Skip to main content

Full text of "Impact of externally set goals on motivation and performance in the sales force"

See other formats










Dedicated to my parents, Kumkum and Bhabatosh Sen Chowdhury, who have 
been my first educators and much, much more . . . 


I gratefully acknowledge the inspiring guidance provided to me by 
Professors Barton A. Weitz and John G. Lynch, Jr. The process of being 
advised on my dissertation was one of the most rewarding experiences I 
have encountered. I am thankful that I had been fortunate enough to be 
able to work under the guidance of Bart. During the course of my 
dissertation, he not only taught me many things that relate to good 
research but also enriched me with knowledge that, I hope, will help me 
be a better person. 

Throughout the course of my entire doctoral program and beyond, my 
wife, Mona, has been a tremendous source of strength and support for me. 
For me, it is impossible to even visualize myself going through the 
entire course without her being there. I do not have words in my 
vocabulary that can do justice to the deep feeling of appreciation I 
feel toward her. 

Finally, I am thankful to my parents, Kumkum and Bhabatosh Sen 
Chowdhury, for providing me with the all their love and support which 
have given me the strength to pursue all that I have done, the vision to 
recognize the worth of my choices, and for instilling in me a sense of 
pride in everything that I do. 







Overview 1 

The Goal Setting Phenomenon 3 

Objectives 5 

Theoretical Perspectives 6 

Summary 8 


Overview 10 

Goal Setting Research 11 

Locke ' s Goal Setting Theory 11 

Contributions of Locke's Theory--The Role of 

Conscious Intentions 12 

The Concept of the Goal — External vs Internal 13 

Implications 15 

Expectancy Theory 16 

Background 16 

The Expectancy-Value Model 17 

The Expectancy Construct 18 

Antecedents of Expectancy 22 

Implications for Goal Setting 25 

Achievement Motivation 27 

Background 27 

"Motive" and "Motivation" 27 

The Principle of Achievement Motivation 2 8 

Implications 30 

The Competitive Outlook 31 

Self-Efficacy 39 

Background 39 

The Construct 41 

Self -Efficacy and Expectancy — Theoretical Issues ....45 
Self-Efficacy and Expectancy — Measurement Issues ....49 
Self-Efficacy and Self-Esteem 52 


Objectives of the Model 54 

Scheme Adopted for Describing the Conceptual Model 55 

The Extended Compliance Path 56 

Overview of the Path 56 

Cognitive Representation of Goal 57 

Effect on Intention — The Compliance Mechanism 59 

Effect of Intention on Performance 62 

Moderating Influences on the Extended Compliance Path 63 

Effect of Goal Specificity 64 

Effect of Extrinsic Rewards 67 

Effect of Expectancy 72 

External Goal Level and Perceived Extrinsic Rewards 74 

Types of Contingent Compensation Plans 75 

External Goal and Intangible Extrinsic Rewards 78 

Moderating Influences on Intangible Rewards 80 

The Goal-Expectancy Pathway 81 

Nature of the Goal-Expectancy Relationship 85 

The Moderating Effect of Self-Ef f icacy 90 

The Intrinsic Motivation Pathway 92 

The Concept of Intrinsic Rewards 92 

The Rationale for Externally Administered 

Rewards 93 

Limitations of the Expected Utility Viewpoint 95 

Relation Between External Goal and Intrinsic 

Rewards 96 

Effect of Intrinsic Rewards on Performance 98 

The Composite Effect of Self-Efficacy 100 

The Static Model — Summary and Implications 101 

Dynamic Aspects of Goal Setting 103 

Proximal and Distal Goals 103 

Dual Effects of Proximal Goals 106 

Implications 108 


The Experimental Paradigm 113 

The Task Setting 113 

Advantages of a Microcomputer Based Experimental 

Methodology 113 

The Self-Efficacy Scale for Negotiation Tasks 118 

The Study 122 

Objective 122 

Subjects 122 

Experimental Design 122 

Experimental Hypotheses 124 

Procedure 127 

Summary 142 


Analyses of the Self-Efficacy Measure 143 

Stability of Self-Efficacy 143 

Effects of Self-Efficacy 146 

Tests of Hypotheses 148 

Performance Levels 148 

Expectancy of Task Success 156 

Intention to Perform and Intrinsic Rewards 159 

The Inverted-U Relationship Between Goal and 

Effort Expended 161 


Implications of Observed Results 163 

Future Research Directions 165 












Abstract of Dissertation Presented to the Graduate School 

of the University of Florida in Partial Fulfillment of the 

Requirements for the Degree of Doctor of Philosophy 




Jhinuk Chowdhury 

August 1990 

Chairman: Professor Barton A. Weitz 
Major Department: Marketing 

An important characteristic of the sales management context is the 
widespread use of and importance attached to sales quotas. The levels 
of sales quotas assigned to salespeople are believed to affect their 
work motivation and actual performance levels. Sales managers, however, 
have little or no recourse to theory-based guidelines that may help them 
in choosing an optimum sales quota in a given situation. 

In the course of addressing issues related to the setting of 
optimal levels of performance targets or goals, this dissertation adopts 
an eclectic perspective and introduces multiple viewpoints in research- 
ing phenomena related to motivation in the sales force. Additionally, 
it proposes an integrating conceptual framework that is used to resolve 
the seemingly contradictory positions of different research traditions 
in related disciplines. The conceptual model proposed in the disserta- 
tion describes the effects of goal setting in terms of intervening 

cognitive and motivational constructs and identifies the relevant inter- 
relationships in order to account for the phenomena associated with goal 

In order to validate some of the key propositions that are genera- 
ted from the conceptual model, an experimental study was conducted. The 
task situation chosen for the study included negotiation sessions in 
which the experimental subjects assumed the role of salespeople and 
participated in bargaining activities with a microcomputer that assumed 
the roles of purchase representatives. A computer program was specifi- 
cally written for this purpose. The dissertation describes in detail 
the methods used for the experimental study and the results of the data 
analyses . 

The results indicate that for every individual, performance in 
sales related tasks improves when the level of difficulty of the goal is 
increased till a certain point. Thereafter, further increases in the 
level of goal difficulty do not improve performance. The results also 
indicate that for low levels of goal difficulty, most individuals 
perform equally. However, when the goals are made increasingly 
difficult, individuals with higher levels of self-efficacy outperform 
those who are low in self-efficacy. 




Within the context of sales management, the practice of setting of 
sales quotas, in one form or another, has been and continues to be 
pervasive. The use of quotas in sales situations is estimated to have 
been used as early as 1888 (Phelps and Westing 1968, p. 739) . In a 
survey conducted well over half a century ago, Griffin (1928) found that 
a majority of the larger manufacturers in the US used some kind of sales 
quota plan. 

With the introduction of the concept of Management by Objectives 
(Drucker 1954; McGregor 1954), interest in the setting of goals or 
quotas was extended beyond the sales force to become a major aspect of 
the organizations' overall development. 1 According to Futrell, the 
concept of Management by Objectives (or MBO) has received "considerable 
attention in management literature and has been characterized as the 
most pervasive idea in management in the last quarter century" (1981, 
p. 134) . However, Futrell (1981) also points out that the idea of MBO 

1 0diorne (1978) provides a review of literature devoted to issues 
concerning Management by Objectives. 

has been interpreted in ways that are markedly different from each other 

and different components of the overall MBO concept have been emphasized 

by the different advocates of the idea. Consequently, in spite of the 

inspiration it has provided and the enthusiasm generated, the concept 

remains amorphous with regard to its capability to generate concrete 

researchable propositions. Steers and Porter (1974) indicate that books 

written on goal setting or quota setting within the purview of MBO are 

numerous, but they typically possess the distinct characteristic of 

"how-to-do-it" manuals that are based on anecdotal wisdom and lack 

formalized theory testing procedures. 

Modern textbooks on sales and marketing management almost 

unanimously emphasize the significance of quotas in sales situations. 

Sales quotas are used for several purposes, among which the most 

important include forecasting, establishing standards of evaluation, and 

motivating salespeople (Anderson, Hair & Bush 1988; Johnson 1976; 

Stanton & Buskirk 1978) . A section on sales quotas in a recent sales 

management text (Anderson et al. 1988) reveals that there exists 

considerable appreciation for the possibility that sales quotas may have 

strong impacts on motivation and performance of salespeople. Anderson 

et al. suggest that 

professional salespeople . . . have special incentives for 
top performance. These sales incentives or motivational 
targets are called quotas and they may take a variety of 
forms. Successful sales managers who skillfully make use of 
sales force quotas can stimulate their salespeople to 
achieve exceptional individual and team performances. 
[1988, p. 354] 

However, despite the common use of quotas, there exists very limited 
research in the marketing literature that is directed toward 
investigating the effects that quotas have on performance. 

The Goal Setting Phenomenon 

The problems that may be associated with the motivational 
properties of goal setting may be appreciated at an intuitive level. 
For, instance, we may conceive of a sales quota that has been set so 
high that it acts as a source of discouragement to the individual. 
Under such conditions, the quota may actually decrease the level at 
which the salesperson performs. On the other hand, if a quota is set 
too low, then the salesperson's employer could be utilizing only a part 
of the individual's potential, and once again extracting less than 
optimal performance from the salesperson. This scenario is an 
illustration of how changes in the levels of an externally set goal 
could affect performance. 

In addition to the effect of the level of goals on performance, 
the manner in which goals are scheduled also affect task motivation and 
performance, especially when goals are repeatedly assigned within a 
period of time. For instance, research suggests that the frequency with 
which goals are scheduled affect task motivation and performance 
(Bandura & Schunk 1981) . Even when the goal levels are same across two 
or more goal conditions, different frequencies of goal setting may 
account for differences in performance levels. 

An useful taxonomic scheme which may be employed in the context of 
goal setting studies involves categorizing the issues into the static 
and the dynamic aspects of the phenomenon. The static representation of 
the phenomenon concerns issues that deal with the mechanisms involving 
cognitive processes and subsequent motivational properties of quotas or 
performance targets at a given point in time (or, at most, over brief 
periods) . It addresses issues that concern the effect of the level of 
externally set goals on motivation and performance in terms of 
intervening cognitive variables, without explicit consideration of any 
effects of repetition and feedback. 

The dynamic aspect of goal setting, on the other hand, concerns 
the issue of how performance goals interact with other factors over 
extended periods of time in its influence on motivation and performance. 
This aspect of goal setting considers how quota levels, performance, and 
feedback at a given point in time affect subsequent performance with 
respect to quotas. From a management perspective, questions related to 
the dynamic aspects of goal setting are at least of equal importance to 
the questions surrounding its static aspects. In a sales management 
context, quotas are important recurring features of business periods. 
Insight into the dynamics of the effect of quotas is, in the very least, 
relevant and perhaps even crucial from the standpoint of understanding 
and analyzing salesforce motivation. 


Although relatively unknown to literature on salesperson 
performance in marketing, goal setting within a broader organizational 
context has been a subject of extensive research in organizational 
behavior. Included among those studies are the few that address the 
effect of goal setting in sales situations (as, for example, Hollenbeck 
& Williams 1987; Huber & Neale 1987) . 

Studies that have explored the effects of goal setting have taken 
into consideration a wide variety of situations and subjects. For 
instance, goal setting studies have inquired into the performance of 
scientists and engineers (Andrews & Farris 1972), productivity of 
factory workers (Dachler & Mobley 1973), performance of chess players 
(Campbell & Ilgen 1976), behavior of households with regard to 
residential energy conservation (Becker 1978), class performance of 
second through fourth grade elementary school students (Hall & Hall 
1976), and behavior of a university hockey team (Anderson, Crowell, 
Doman & Howard 1988), among several other application domains. For a 
comprehensive review of goal setting studies, refer to Locke, Shaw, 
Saari & Latham (1981) . Tubbs (1986) reports results of a meta-analysis 
of 87 studies in the general area of goal setting for the purpose of 
estimating the extent of empirical support for the major postulates of 
Goal Setting Theory. 

Despite the rather impressive list of goal setting studies in 
organizational behavior, with a few notable exceptions, most of the 

studdies have focussed on amassing empirical descriptions of the goal 
setting effect. Inquiries into the motivational and cognitive 
mechanisms that may account the phenomenon have been unsystematic and 
rare. Partly as a consequence, goal setting literature has been beset 
by conflicts among researchers of different paradigms and paradoxical 
observations. The objectives of this dissertation include addressing 
the substantive issues related to goal setting and at the same time 
proposing an overarching conceptual framework that may account for the 
seemingly conflicting positions of different research traditions. In 
addition to providing an integrating conceptual framework, this 
dissertation addresses the contradictory findings in previous studies 
and develops an experimental methodology for testing the theoretical 
propositions . 

The conceptual model describes the goal setting process in terms 
of intervening cognitive and motivational variables and identifies the 
relevant interrelationships in order to account for the static aspects 
of the goal setting phenomenon. Subsequently, the mechanisms and 
variables identified in the conceptual framework are used to explore 
some of the dynamic aspects of goal setting. There exists very little 
research that has been directed toward these issues. 

Theoretical Perspectives 

In the marketing literature, an overwhelming proportion of prior 
research on salesforce motivation has adopted the Expectancy-Value 
framework (e.g., Cron, Dubinsky & Michaels 1988; Ingram & Bellenger 

1983; Oliver 1974; Teas 1981; Tyagi 1982; and Walker et al . 1977) . 
Predictions generated from the Expectancy-Value model can help in 
explaining the motivational effects of externally set goals on 
salespersons' performance. However, there exist other paradigms that 
can be used to examine this issue and have implications that sharply 
oppose the implications of Expectancy-Value Theory. 

Since this dissertation actually addresses a phenomenon rather 
than a specific question of theoretical interest alone, the theoretical 
perspective adopted is eclectic rather than exclusive. In addition to 
Expectancy-Value Theory (Vroom 1964, 1966) and Locke's Goal Setting 
Theory (Locke 1966, 1968), references are made to Atkinson's Achievement 
Motivation Model (Atkinson 1964, 1978; Atkinson & Feather 1966) . 
Additionally, concepts drawn from Social Learning Theory (Bandura 1977a 
& b, 1982a & b; Bandura, Adams & Beyer 1977) are introduced to develop 
an integrative framework. 

The framework presented suggests that these different theoretical 
perspectives are not necessarily opposed to each other as has been 
traditionally viewed in organizational behavior literature. On the 
contrary, these perspectives address separate although partly 
overlapping domains that can be made to mesh together in an overarching 
framework, so long as they are correctly interpreted. 

This integrative theoretical perspective lends a richer accounting 
of the overall goal setting phenomenon, contributes to our understanding 
of substantive issues that had previously remained obscure, and provides 

a framework based upon which subsequent research explorations may be 


An important characteristic of the sales management context is the 
widespread use of and importance attached to sales quotas. These 
performance targets are also recognized to influence motivation and 
performance in the sales force. In marketing literature dealing with 
sales force issues, there exists very little prior research on the 
motivational effects of performance goals and related concerns. 

In organizational behavior literature, on the other hand, research 
on goal setting has a rich tradition. However, limitations of prior 
research in this area include fragmented treatments of the processes 
involved in the goal setting effect. A large portion of the work done 
has been devoted to obtaining empirical support for select propositions 
within the paradigm, while attempts to explore the mechanisms and 
processes responsible for the goal setting phenomenon have been rare and 
fraught with controversies. 

This dissertation contributes to the literature on salespersons' 
motivation and performance in marketing by 

(1) introducing emerging perspectives to the area; specifically, Locke's 
Goal Setting Theory, Atkinson's Achievement Model, and Bandura ' s concept 
of self-efficacy; 

(2) developing an integrative conceptual framework that reconciles 
conflicting theoretical positions; 

(3) empirically examining the interrelationships among the various 
factors that are involved in the static aspects of the goal setting 
process; and 

(4) extending prior research by examining issues related to the dynamic 
aspects of the effects of quotas on motivation and performance. 

While the research in this dissertation is undertaken in a sales 
force context, the framework, propositions, and empirical results have 
broader implications. Estimating market share goals or objective arise 
in many marketing management and strategic management contexts. The 
proposed framework can potentially be modified to suit any 
organizational context within which motivated performance is important. 
The level at which the theoretical constructs have been dealt with in 
this thesis lends a good measure of generality to the propositions and 
recommendations outlined later. 



The integrated framework outlined in the next chapter draws upon 
concepts in Vroom's Expectancy-Value Model, Locke's Goal Setting Theory, 
Atkinson's Achievement Motivation Model, and Bandura's Self-Efficacy 
Theory. In this chapter, each of these paradigms is briefly reviewed 
and critically evaluated in the context of the goal setting phenomenon. 

In organizational behavior literature, Vroom's Expectancy-Value 
Theory, Atkinson's Achievement Motivation Theory, and Locke's Theory of 
Goal Setting have been regarded for the most part as rival paradigms 
that offer competing recommendations regarding how goals should be set 
in order to maximize performance (Erez & Zidon 1984; Garland 1984; 
Hollenbeck & Klein 1987; Janz 1982; Matsui, Okada & Mizuguchi 1981; 
Mento, Cartledge & Locke 1980; Motowidlo, Loehr & Dunnette 1978) . Some 
researchers have maintained that these research streams originate from 
separate traditions and that the differences between them are 
irreconcilable (e.g., Mento et al . 1980); while others have acknowledged 
the conflicts but have attempted to account for the predictions of one 
theory in light of another viewpoint (e.g., Matsui et al . 1981). 

Following the overview, the remainder of this chapter is divided 
into five more sections. Sections two, three, and four broadly outline 



each of Locke's, Vroom's, and Atkinson's model in turn, and describe how 
the proposals of each have been typically interpreted in the context of 
goal setting. The fifth section of the chapter examines in somewhat 
more detail studies dealing with and issues surrounding the comparisons 
and contrasts between these three theoretical viewpoints. The sixth and 
final section of this chapter is a critical exposition of the self- 
efficacy construct and its potential for contribution to the domain of 
goal setting. 

Goal Setting Research 

Locke' s Goal Setting Theory 

Locke's hypotheses concerning the effect of goals on performance 
have generally been referred to as Locke's Goal Setting Theory. The 
postulates of Goal Setting Theory are: 

(1) hard goals produce a higher level of performance 
(output) than easy goals; (2) specific hard goals produce a 

higher level of output than a goal of "do your best"; and 
(3) behavioral intentions regulate choice behavior. [Locke 

1968, p. 157] 

Locke et al . (1981) clarify that a basic assumption of the goal 
setting hypotheses is that the assigned goals are accepted by the 

individual. In other words, the goal setting effect (difficult and 
specific goals contributing to improved performance) may be expected to 
occur only when goals are accepted by the individuals to whom the goals 
are assigned. 


Contributions of Locke's Theory — The Role of Conscious Intentions 

Among the earliest instances of a laboratory study inquiring into 
the impact of goals on performance is included the work of Mace (1935) , 
who found that subjects assigned a higher performance goal on a 
psychomotor task worked at a faster pace than those assigned a lower 
goal. Locke and his associates have played a major role in developing 
this finding into a stream of research on goal setting. Based on their 
investigations of task goals and motivated performance, they have 
established an influential research paradigm in organizational behavior 
that builds upon Ryan's (1958) conceptualization of motivated behavior. 
According to Ryan, one very important but under-emphasized determinant 
of motivated performance is the individual's intentions . 

Previous research on issues related to behavior could be 
classified into one of two polarized groups. Behaviorists in the 
tradition of Skinner 1 chose to explain behavior without any reference 
whatsoever to the individuals' internal mediating processes; while most 
motivation and personality psychologists exploring the phenomena of 
performance emphasized processes that have been either left inexplicated 
or assumed to involve complex unconscious reasoning on part of the 
individual . 

Locke (1966, 1968) highlights Ryan's notion that it is perhaps 
futile to attempt to conduct a laboratory experiment involving different 

x Refer to Skinner (1953, 1971, 1974) for detailed expositions of 
his position on the proper mode of inquiry into behavioral phenomena. 


levels of performance of human subjects, in which the conscious 
intentions are not manipulated. Locke (1966) states that in electing to 
focus on uncovering the effects of the level of intended achievement on 
the actual level of achievement, his line of research departs sharply 
from the earlier, well-known approaches inquiring into the phenomenon of 
motivated performance. In his opinion, previous research on the 
determinants of performance, such as those that consider the effect of 
several social, situational or psychological factors on performance, do 
not specify what role the individual's conscious intentions play in the 
process and consequently remain considerably ambiguous. Locke (1966) 
mentions, for instance, the work of Atkinson (1958) and McClelland, 
Atkinson, Clark & Lowell (1953) on the relationship between need for 
achievement and performance. These researchers specifically acknowledge 
that the influence of "the need for achievement" on performance does not 
occur through the individual's conscious experience, even though it 
affects behavior visibly. 

The Conce pt of the Goal — External vs Internal 

There have been some confusion regarding what goals actually mean 
due to imprecise definitions in some cases and differences in the 
theoretical foci in others. For an individual, goals may be externally 
set (i.e., set by another person or agency) or goals may be thought of 
as having been self-set. There exist strong theoretical and empirical 
justifications in favor of the argument that these two types of goals 
are substantively very different in terms of the way they influence 


performance. While some later researchers have consciously and 
deliberately explicated what they mean by their use of the term goal 
with regard to whether it is externally set or self -set (e.g, Garland 
1983, 1985) , a number of previous studies suffer from ambiguity 
concerning this issue. 

Locke et al . define a goal as "what an individual is trying to 
accomplish; it is the object or aim of action" (1981, p. 126) . Based on 
this definition, it would appear that they are referring to an internal 
or self -set goal. This definition evidently obviates the issue of goal 
acceptance or goal rejection (by the individual), as is characteristic 
of self-set goals. However, they also acknowledge other frequently used 
concepts that are similar in meaning to that of a goal, as for instance, 
performance standard, quota, work norm, objective, deadline, and budget. 
Additionally, in their earlier experimental paradigm as well as in their 
statement of hypotheses, goals are, for the most part, identified with 
specific externally assigned target performance levels. 

Locke's initial imprecision with regard to the the distinction 
between internal and external goals can be understood when one considers 
that the need for this distinction had not arisen at the time, and that 
Locke and his associates wanted to project a cognitive-psychological 
orientation as opposed to a stimulus-response behaviorist approach. 
Nevertheless, their enunciation of the construct has been subject to 
questioning, if not criticism. According to Naylor & Ilgen, "in order 
to describe goals with any degree of precision, much more must be 
addressed beyond {their} definition" (1984, p. 97) . Garland similarly 


points out that "confusion in literature . . . may stem from a lack of 
clear distinction between an externally assigned goal and a task goal, 
as defined here. . . . The task goal is self-set; thus, it makes little 
sense to question its acceptance by the individual" (1985, p. 348) . 

As the postulates of the Goal Setting Theory indicate, most of the 
early work and a substantial proportion of contemporary goal setting 
research concern externally set goals. Garland (1985), on the other 
hand, is an instance of a study that inquires into questions related to 
self -set or "task" goals. Research on externally set goals and self -set 
goals are not incompatible, however; a particular study may choose to 
include both kinds of goals as its objects of inquiry. As long as the 
concepts are clarified at the outset, it should not create either 
conceptual or communication problems. Garland (1983) and Stedry (1960) 
are examples of studies that include both assigned goals (external) and 
personal goals (internal) in their ambits. 


In espousing his theoretical viewpoint and providing rationale for 
his hypothesis, Locke (1966, 1968) and his associates (Locke et al . 
1981) explicitly indicate that their conceptualization poses a challenge 
and an alternative to Vroom's and Atkinson's notion of how motivated 
performance occurs. In their opinion, the idea that motivation and 
performance can be correlated with expectancy of task success not only 
contradicts their viewpoint but is flawed as well. The body of 


empirical research in the tradition of Locke thus emerged as a rival 
paradigm to the theoretical perspectives of Vroom's and Atkinson. 

In organizational behavior literature, however, an overwhelmingly 
large proportion of studies report only empirical support for the goal 
setting hypotheses without attempting to inquire into its theoretical 
structure. Consequently, goal setting has acquired the distinct flavor 
of being an applied tool to a greater extent than the reputation for 
being a cognitive construct at the focus of rigorous theoretical 
inquiry. It may be worthwhile to note at this juncture that some 
researchers have referred to the domain of goal setting variously as a 
"motivational technique" (Latham & Locke 1979; Locke & Latham 1984), a 
"motivational technology" (Naylor & Ilgen 1984), and the "technique of 
goal setting" (Locke 1978) . 

The framework proposed in this study attempts to account for the 
cognitive mechanisms responsible for the goal setting effect and 
interrelates them with the Expectancy Theory constructs. 

Expectancy Theory 


The Expectancy-Value Model of motivation, first introduced to 
organizational behavior by Vroom (1964), has had a substantial impact on 
research on salespersons' motivation in the marketing literature. 
Oliver (1974) first applied the Expectancy-Value Model, also referred to 
as Expectancy Theory, in a sales context. Walker et al . (1977) followed 
soon after and offered a conceptual model that identified a set of 


individual, interpersonal, organizational and environmental variables 
that are likely to influence a salesperson's motivation and performance. 
While their model incorporates many concepts and research findings from 
industrial psychology, the foundation of the model rests on Expectancy 
Theory. Subsequent studies that have adopted the Expectancy -Value Model 
in some form or another include Cron, Dubinsky & Michaels (1988); Ingram 
& Bellenger (1983); Teas (1981); Teas & McElroy (1986); and Tyagi (1982, 
1985) . 

Oliver (1974) notes that Expectancy Theory has its origins in the 
pioneering works of Lewin (1938) and Tolman (1932) . Their studies 
suggest that humans (among several other organisms) possess approach- 
avoidance tendencies, which are influenced by cognitive representations 
of desirabilities of the end states of their target actions as well as 
some kind of mental calculus reflecting the contingency of the target 
action and the end state. 

The Expectancy-Value Moriffl 

The basic postulate of Expectancy Theory as applied to 
organizational behavior (Vroom 1964) is that an individual's job 
performance (P) is a multiplicative function of his/her motivation (M) 
and ability (A), i.e., 

P = / (M x A) . 
Motivation (M) , in turn, is a multiplicative function of expectancy (E) , 
valence (V), and instrumentality (I), so that 

M = / (E x V x I) . 

Expectancy refers to the individual's subjective probability that effort 
will lead to a given level of performance, while instrumentality (I) is 
a subjective correlation 1 that the same performance level will be 
associated with certain outcomes. The valence (V) refers to the 
desirabilities of each of the outcomes associated with the given 
performance level. Valences can be both positive and negative. 

One implication of Expectancy Theory which has wielded 
considerable influence on subsequent theorizing is the idea that the 
desirability of a reward and the individual's estimate of how likely he 
or she will be in securing the reward constitute equally important 
motivators of action. 

The Expectancy Construct 

Within the Expectancy-Value framework, the expectancy construct 
refers to an individual's assessment of the contingency between the 
level of effort expended at specific tasks and being able to achieve a 
specific level of performance. In the context of salespersons' 
motivation, expectancy may be defined as 

instrumentality, unlike expectancy, is not a probability 
estimate. Expectancies are typically scaled to range from to 1.0, 
while the values of instrumentality are scaled to range from -1.0 to 
+1.0. Instrumentality is defined as the extent to which a given 
performance level is estimated to cause the attainment of an outcome or 
lead to the blocking (prevention) of the outcome. Mitchell (1979) 
argues that these properties of instrumentality make it a correlation 
estimate rather than a probability estimate. He notes, however, that 
several studies have failed to notice this subtlety and have measured 
instrumentality incorrectly. 


the salesman's estimate of the probability that expending a 
given amount of effort on task (i) will lead to an improved 
level of performance on some performance dimension (j) . 
[Walker et al . 1977, p. 162] 

This conceptualization of expectancy adopted by Walker et al . (1977) is 
actually similar to a refinement of the construct of expectancy that was 
originally proposed by Porter & Lawler (1968) and subsequently amplified 
by Campbell, Dunnette, Lawler & Weick (1970) . These researchers were 
instrumental in dividing expectancy into two parts: E-I, the perceived 
contingency between effort and performance , and E-II which refers to the 
perceived contingency between performance and outcomes . Prior to this, 
expectancy was often used to refer to the perceived contingency between 
effort and Outcome. It is the E-I representation of expectancy that is 
reflected in the definition of Walker et al . (1977) and the one that is 
more commonly accepted in organizational behavior literature. While 
E-II also measures some aspect of expectancy, that measure is viewed as 
exerting no or, in the very least, an insignificant amount of impact in 
comparison with E-I. Expectancy I or E-I measures how the individual 
evaluates the possibility of task success : while E-II is, in a sense, a 
probabilistic estimate of the worth of that task success. That is, E-I 
is visualized as being more contiguous to the self and the motivational 
systems of the individual than any other measure. This idea is also 
explicitly advocated in Bandura's Self-efficacy Theory (Bandura 1977a & 
b, 1982a & b) as will be discussed in detail later. 

Many researchers who have dealt with the Expectancy-Value 
framework, including some in applied psychology, have been quite 
unconcerned about this point regarding the concept of expectancy and 


have interpreted the construct too broadly. Expectancy is often used as 
a global term to refer to the subjective probability that action will 
lead to a set of outcomes. These outcomes have been specified at very 
general levels and do not correspond to specific performance levels 
suggested in the refinements of the Expectancy-Value framework. 

The misconceptions regarding what the expectancy construct 
actually stands for is evident in the way it has been measured in a 
number of studies. For example, Oliver (1974) operationalized 
expectancy as the subjective probability of an individual that his or 
her efforts would lead to the attainment of performance goals. The 
expectancy construct, as operationalized by Tyagi (1982, 1985), measures 
individuals' probability estimates that working hard will lead to high 
productivity, good job performance, and timely completion of work. In 
one of the two measures of expectancy, Kohli (1985) used a 3-item 
instrument that measured individuals' perceptions of "linkages between 
working hard and productivity, job performance, and doing a job well" 
(p. 428) - 1 

Sujan (1983) draws attention to this discord between the 
operational and conceptual definitions of expectancy. He speculates 
that the difficulty in measuring expectancy at the specific level could 

1 In another measure of expectancy, however, Kohli (1985) used a 
modified version of a scale originally used by Sims, Szilagyi & McKemey 
(1976) . The original scale was designed to measure "Expectancy I" (or 
E-I), as proposed by Porter & Lawler (1968) and subsequently amplified 
by Campbell, Dunnette, Lawler & Weick (1970) . 


have forced researchers of salesforce motivation to adopt more easily 
assessed global measures. He suggests that for a salesperson with 
several customers, a proper expectancy measure would involve questioning 
the salesperson (separately for each customer) regarding his/her 
estimates of the chances of achieving a target level of sales, given 
that he/she spent only a specific level of effort on the customer. 
Additionally, such measures should also be obtained for several 
different levels of effort (from 50% below normal to 50% above normal, 
for instance) in order to have a reasonable understanding of the 
salesperson's perceived effort -performance relationship. For example, 
one measure of expectancy used by Teas (1981) did consist of probability 
estimates concerning the the productivity of a 10% increase in selling 
efforts, a 10% increase in time devoted to obtaining to new accounts, 
and a 10% increase in time devoted to selling activities. 

Another method similar to the one suggested by Sujan (1983) and 
adopted by Teas (1981) could involve obtaining separate expectancy 
measures corresponding to different performance levels (or goals) . That 
is, the questions used for measuring expectancy could probe for 
individuals' probability estimates of achieving different levels of 
performance standards, instead of varying the projected effort levels. 
In fact, as will be discussed in more detail later, this 
conceptualization of expectancy is the one that is most meaningful 
within the context of goal setting research. 


Antecedents of Expsntanr.y 

Uncovering factors that contribute to the development of 
expectancies has been an important objective of several studies (e.g., 
Oldham 1976; Peters 1977; Pritchard, De Leo & Von Bergen 1976; Sims, 
Szilagyi & McKemey 1976 in organizational behavior and Kohli 1985; Sujan 
1983; Teas 1981; Teas & McElroy 1986; Tyagi 1982 in salesforce 
motivation literature). Sujan (1983), for instance, summarizes the 
antecedents of expectancy and categorizes them into one of the following 
four classes: ability related (such as, self-esteem, skills, and 
experience) , reward contingencies (perceived link between effort and 
obtaining rewards), task related (such as, competition, territory 
potential, general economic condition, etc.), and omanizati nnal factors 
(e.g., job challenge and variety, job importance, task conflict, role 
overload, etc. ) . 

For the purpose of this research on goal setting, a simpler, 
dichotomous classification scheme suffices. This scheme categorizes the 
antecedents of expectancy into either an internal or an external group. 
As a cognitive variable, expectancy is typically hypothesized to be 
influenced by other cognitive factors. That is, the probability 
estimates that constitute expectancies are viewed as consequences (or 
even integrations) of other cognitions. These two broad categories 
relate to Sujan' s (1983) four categories as follows: internal factors , 
involving cognitions related to the self, and external (or task) factors 
that primarily involve the individual's perceptions of job 
characteristics and job environment. While the results of much of prior 


research on the antecedents of expectancy are significant for goal 
setting research, the proposed internal/external categorization is the 
one that contributes most directly yet parsimoniously to the 
understanding of the forthcoming frameworks. 

Internal antecedents of expectancy . in a review of organizational 
behavior research related to antecedents of expectancy, Mitchell (1979) 
identifies the following studies that considerably enrich our knowledge 
of internal (or self -related) factors which potentially contribute to 
expectancy. Oldham (1976) observed that individuals with higher self- 
esteem also possessed higher expectancies. A number of studies (Lied & 
Pritchard 1976; Mitchell, Smyser & Weed 1975; Sims, Szilagyi & McKemey 
1976) found that individuals with internal locus of control have higher 
expectancies than those with external locus of control. The 
interpretation of this research suggested by Mitchell is that 

people who are confident about their ability to influence 
the world around them see stronger relationships between 
what they do and the results of their actions than do people 
who see these outcomes occurring as a function of fate or 
luck. [1979, p. 255] 

In literature on salesforce motivation, Walker et al . (1977) include 

self-esteem and perceived competence among suggested antecedents of 

expectancy. Their rationale is based upon the earlier work of Korman 

(1971) and Lawler (1970) . Consistent with the preceding observations, 

Kohli (1985) found that salespeople's self-esteem has a strong effect on 

their expectancies. Teas (1981) found specific self-esteem to be 

positively related to expectancy. Earlier, Bagozzi (1978, 1980) had 

observed that self-esteem was related to performance. 


In sum, there exists strong theoretical as well as empirical 
support for the contention that cognitions related to the self are 
included among valid antecedents of expectancy. 

External antec edents of expectancy . Among external factors that 
are theorized to affect expectancy are included general economic 
conditions, territorial potential, competition and raw material 
shortages (Walker et al . 1977). Tyagi (1982) provides evidence that 
expectancy is also affected by "organizational climate variables" such 
as job challenge and variety, task conflict, job importance, leadership 
consideration and other such factors. 

In the context of goal setting research, it is important to 
realize that along with perceptions concerning the self, task 
characteristics (or perceptions thereof) influence expectancy estimates 
as well. It is intuitive and completely reasonable to assume that as 
the task actually gets more difficult, individuals' perceptions of task 
difficulty will also reflect that change to a great extent. In other 
words, individuals' perceptions of task difficulty (and therefore 
expectancy estimates) are expected to follow and be influenced by 
objective characteristics of the task, including level of difficulty. 

In organizational behavior literature, in an article reporting the 
results of an empirical study on the relationship between task design 
and individual responses, Alegra (1983) shows that an individual's 
perceptions reflect the objective character of the task and influence 


attitudinal and behavioral responses. 1 Within the realm of goal setting 
research, this translates into the idea that "as goals difficulty 
increases, the {subjective} probability of goal attainment decreases" 
(Campbell 1982, p. 80) . The idea that more difficult goals contribute 
to lowered subjective probabilities (or expectancies) of task success 
has been widely tested and accepted in this research domain (e.g, 
Garland 1982, 1983, 1984; Locke 1964; Matsui, Okada & Mizuguchi 1981; 
Mento, Cartledge & Locke 1980; Motowidlo, Loehr & Dunnette 1978) . 

Implic ations for Goal Setting 

Since expectancy of task success may be expected to diminish as 
the externally set goal level is increased (i.e., made more difficult), 
Vroom's Expectancy Theory has often been interpreted to imply that as 
goals are made more difficult, motivation and task performance decrease 
monotonically (Garland 1984; Janz 1982; Motowidlo et al . 1978; Mento et 
al. 1980); or conversely, as subjective probability of success (or 
expectancy) increases, performance improves correspondingly (Dachler & 
Mobley 1973; Matsui & Tashitake 1975; Schwab & Dyer 1973; Sheridan, 
Slocum & Min 1975) . 

1 This study is a methodological refinement of a line of research 
based on the "job characteristic approach" (Hackman & Lawler 1971; 
Hackman & Oldham 1975, 1976) . This approach views the task as perceived 
by the task performer as the determinant of behavior and attitude. 
Included among the conclusions arrived at through these studies is the 
one that suggests that perception of task performers is reasonably well 
grounded in objective reality. 


Matsui et al . (1981), however, propose a more sophisticated 
explanation of how Expectancy Theory postulates may account for Goal 
Theory predictions. They suggest that for easy goals, while the 
expectancies for reaching the goal are high, the valences of doing so 
may be low. For difficult goals, the values of expectancy and valence 
would be reversed. It should be noted that treatment of valence in 
Matsui et al . (1981) is different from the traditional Vroomian 
exposition of the construct. In Vroom's theory, valence of an outcome 
is a summation of the valences (or the desirabilities) of the rewards 
associated with the outcomes. The conceptualization proposed by Matsui 
et al. (1981), on the other hand, bears a great deal of resemblance with 
Atkinson's notion of the incentive value of task success. Even so, the 
notion that difficult goals may be valued differently than easy goals is 
an important one in this context. 

Garland (1984) contends that manipulating expectancy by varying 
goal difficulty violates the boundary conditions of Expectancy Theory. 
In his opinion, support for Expectancy Theory predictions may be found 
within a goal setting experiment provided certain conditions are met. 
Details of these and other issues will be discussed in a forthcoming 
section of this chapter devoted to comparisons and contrasts between the 
models . 

In sum, however, theories originating from the expectancy-value 
tradition, including Atkinson's model, have for long been considered to 
be in conflict with Locke's proposals. 

Achievement Motivation 


The Theory of Achievement Motivation (Atkinson 1964, 1978; 
Atkinson & Feather 1966) shares its genesis with Expectancy Theory in 
the early works of Lewin and his associates on the resultant valence 
theory of the level of aspiration (refer to Lewin, Dembo, Festinger, & 
Sears 1944, for a summary), and in the works of Tolman (1932, 1955). 
Consequently, it bears some resemblance with Expectancy Theory in its 
structure. At the same time, however, the Theory of Achievement 
Motivation possesses certain distinctive characteristics. 

This line of research evolved from the works of a team of 
researchers which began at the University of Michigan in the late 1940s 
in the studies of the effects of hunger, and later of experimentally 
induced motivation to achieve, on the content of thematic apperception, 
i.e., imaginative behavior (Atkinson & McClelland 1948; McClelland & 
Atkinson 1948; McClelland, Clark, Roby & Atkinson 1949) . Achievement 
Motivation Theory essentially developed from a research stream with a 
strong interest in enduring individual differences and their effects on 
performance. In contrast, Expectancy Theory centered more on the 
subjective analyses of potential rewards arising from behaviors. 

"Motive" and "Motivation" 

Within the framework of Achievement Motivation, the terms "motive" 
and "motivation" have distinct meanings. Motive usually refers to an 
individual difference variable, an inherent and stable property of the 


personality. It is often characterized as "a latent disposition to 

strive for a particular goal-state or aim, e.g., achievement, 

affiliation, power" (Atkinson & Reitman 1958, p. 279). Atkinson's 

formal definition is as follows. 

The term, motive , has been used to refer to the disposition 
within the person to strive to approach a certain class of 
positive incentives (goals) or avoid a certain class of 
negative incentives (threats) . [1958, p. 303] 

Motivation, on the other hand, is a term used to refer to the 

tendency, readiness, or inclination to initiate behavior. It has been 

characterized as the "aroused state of the person that exists when a 

motive has been engaged by the appropriate expectancy" (Atkinson & 

Reitman 1968, p. 279) . This use of the term, motivation, is very 

similar to, if not identical with, the usage in most psychological 

theories, including Expectancy Theory. In Atkinson's language, 

the arousal of motivation to approach, i.e., to perform the 
act, is equivalent to the expected positive utility of the 
consequences. Here the term motivation is used to designate 
the activated state of the person which occurs when the cues 
of a situation arouse the expectancy that performance of an 
act will lead to an incentive for which he has a motive. . . 
. The arousal of motivation to avoid, i.e., not to perform 
the act, is equivalent to the expected negative utility of 
the consequences. . . . The resultant motivation, which is 
expressed directly in performance, is a summation of 
motivation to approach and motivation to avoid. [1958, 
p. 304] 

The Pr inciple of Achievement Motivation 

In mathematical terms, Atkinson suggests that for an individual 
the tendency to achieve success, T s , can be expressed as follows. 

T s = m s • p s • Isi where, 
m s = the "motive" to achieve success, 


P s = the subjective probability (expectancy) of 

achieving success, and 

I s = the incentive value of success. 
A crucial assumption in Atkinson's formulation of motivation is 
that the incentive value of achieving success is related negatively to 
the expectancy of achieving success as follows: I s = (1 - P s ). This 
negative relationship implies that the incentive value of success, i.e., 
the satisfaction derived from achieving the goal, is lowest for 
extremely easy tasks (which have high expectancies) and is the highest 
for very difficult or impossible tasks (whose expectancies are close to 
zero) . Consequently, the expression for T s reduces to 
T s = m s . (P s ) . (1 - p s ) . 

Similarly, the tendency of the individual to avoid failure, Tf, 
can be expressed as Tf = mf . Pf . If, where, 
ffif = the "motive" to avoid failure, 
Pf = the subjective probability of failure, and 
If = the incentive value of failure. 
Again, it is assumed that the incentive value of failure , If, is 
equal in magnitude but opposite in sign to the subjective probability 
(or expectancy) of success , P s . In other words, If = (- p s ). This 
implies that the incentive value of failure is always a negative 
quantity (or at best zero), since failure is a noxious event. Also, if 
the possibility of success is high (i.e., P s is high), then the act of 
failing is more negatively evaluated than the converse situation where 


the probability of success is low. The expression for Tf then reduces 
to T f = mf .Pf. (- P s ) . 

Finally, Atkinson assumes that P s + Pf = 1; so that Pf = (1 - Ps) ■ 
Consequently, we have Tf = mf . (1 - P s ) . (- P s ) ; 
i.e, T f = = - mf . (P s ) . (1 - P s ) . 

The total (or resultant) tendency to approach or avoid, i.e., the 
consequent motivation, T = T s + Tf. 

That is, T = [m s . (P s ) . (1 - P s ) ] + [- m f . (P s ) . (1 - P s ) ] , 
or T = (m s - mf) . (P s ) . (1 - P s ) . 

If "m" represents the difference between the motive to achieve 
success and the motive to avoid failure, i.e., if m = (m s - mf ) , then 
the resultant tendency or motivation, T, may be expressed as 
T = m. (P s ) . (1 - P s ) , or, alternatively as 
T = m. [P s - (P s ) 2 ] . 


The preceding mathematical expression for T suggests that the 
value of T is a continuous function of P s . For positive values of m, T 
is minimized when P s = or P s = 1.0; and maximized when P s = 0.5. That 
is, for individuals in whom the motive to achieve success exceeds the 
motive to avoid failure, the task motivation will be nil when the 
expectancy of task success is or 1; while the motivation will be the 
greatest for tasks associated with a 0.5 subjective probability of 
success. In other words, these individuals are motivated mostly by by 
tasks of intermediate difficulty and their motivation diminishes as the 


perceived difficulty approaches the very easy (complete certitude of 
success) or the impossible (absolute certainty of failure) . 

For negative values of m, the situation is reversed: T is maximum 
at P s = and P s = 1.0; and minimum at P s = 0.5. That is, for indi- 
viduals in whom the motive to avoid failure exceeds the motive to 
achieve success, task motivation will be minimum for tasks of 
intermediate difficulty and maximum for tasks that are either very easy 
(complete certitude of success) or impossible (absolute certainty of 
failure) . 

For individuals in whom the two motives are equal, task motivation 
is not contingent upon expectancy or subjective probability of success. 

The Competitive Outlook 

This section discusses the several studies that address the mutual 
conflicts and inconsistencies embodied in Locke's, Vroom's, and 
Atkinson's models. 

Motowidlo et al . (1978) was one of the first influential studies 
to explore the possible contradictions. According to these researchers, 
the three aforementioned theories differ considerably with regard to 
their predictions of how increased levels of externally set goals affect 
performance. Their interpretation of the goal setting effect from the 
point of view of each of the three theoretically distinct positions is 
summarized as follows. 

Vroom's Expectancy-Value Theory proposes that performance is 
directly related to motivation which in turn is a function of expectancy 


and valence. Since difficult goals should translate into lowered 
expectancies, raising the level of difficulty of an externally set 
should diminish motivation and, consequently, performance in the task. 
In other words, Expectancy Theory is interpreted by Motowidlo et al . 
(1978) as suggesting that performance is a monotonically decreasing 
function of the level of difficulty of externally set goals. 

In contrast, Locke's Goal Setting Theory explicitly states that 
more difficult and specific goals result in improved performance, 
provided the externally set goals are accepted by the individual 
performing the task. Ths implies that performance is a monotonically 
increasing function of the level of difficulty of externally set goals. 

Atkinson's Achievement Motivation Theory provides the third 
distinct position with regard to the impact of goal difficulty on 
performance. According to Motowidlo et al . (1978), it follows from the 
propositions of Achievement Motivation Theory that individuals are 
likely to have the highest motivation for (and perform best at) tasks of 
intermediate difficulty. Thus, the relationship between the level of 
goal difficulty and performance is non-monotonic . Specifically, the 
relationship is an inverted U-shaped function, with the highest 
performance occurring at intermediate levels of goal difficulty for 
individuals in whom the motive to achieve success exceeds the motive to 
avoid failure. For individuals in whom the motive to avoid failure is 
stronger than the motive to achieve success, the relationship is an U- 
shaped function with performance at maximum for intermediate levels of 
task difficulty. Finally, for individuals in whom the two "motives" are 


equal, performance should be independent of the level of task 

Motowidlo et al . (1978) suggest that the differential treatment of 
"goal specificity" in the various research paradigms may partly explain 
the fact that scholars of each of these conflicting viewpoints usually 
find support for only their own research hypotheses. For example, Locke 
and his associates have typically focussed on the effects of specific 
goals; while researchers in the tradition of Vroom and Atkinson have 
generally not utilized specific quantitative goals. Instead, task 
difficulty and probability of success were manipulated by reporting to 
the subjects varying proportions of task performers (subjects) who could 
be expected to succeed at the task. Motowidlo et al . (1978) hypothesize 
that for specific goals, task performance and probability of success 
should be negatively related (in accordance with Locke's proposal); 
while for ambiguous goals the relationship should be either positive or 
curvilinear. The authors found no support for Locke's model which 
posits an inverse relationship between expectancy and performance. On 
the contrary, they observed an inverted U-shaped relationship between 
objective probability of task success and performance, with the maximum 
performance occurring at intermediate probability levels. The 
relationship between subjective probability of success and performance 
was found to be linear and positive. 

Mento et al. (1980) interpret the differences in the three 
paradigms similarly and also identify the different research traditions 
from which these separate positions evolved. These researchers, 


however, challenged the validity of the findings of Motowidlo et al . 
(1980) and outlined a number of conceptual and methodological drawbacks 
in that study. According to Mento et al . (1980), the lack of support 
for Locke's assertion (that the expectancy-performance relationship is 
negative and linear) could be attributed to these drawbacks. 

In the first of two experiments reported in Mento et al . (1980), 
the independent variables were goal difficulty and valence (manipulated 
by offering three levels of monetary incentives for reaching the 
assigned goal) . These factors were used in a crossed factorial design 
and their effects on effort and performance were measured. The authors 
observed that performance was unaffected by any Goal Theory or 
Expectancy Theory variables. The results also indicated that subjective 
probability of success (expectancy) was negatively correlated with 
effort at a significant level. However, when the effect of "subjective 
goal difficulty" was partialled out, this correlation was reduced to 
nonsignif icance . The most immediate determinant of effort was the level 
of the individual's personal goal. With this variable partialled out, 
the correlation between subjective goal difficulty and effort reduced to 
nonsignif icance . A supplementary analysis revealed that the personal 
goal effect was greater for subjects with high expectancies than for 
those with low expectancies. 

In the second experiment, the authors attempted to manipulate 
subjective probability of success and goal difficulty, orthogonally. It 
was found that expectancy and valence "were unrelated to effort and 
performance when the other factors were controlled" (Mento et al . 1980, 


p. 438) . However, these authors failed to replicate the negative, 
linear relationship between objective probability of success and 
performance as posited by Locke. 

Garland (1984) attempts to integrate the diverging opinions and 
observations in the two aforementioned studies. He notes that while 
Mento et al . (1980) suggested that Expectancy Theory variables 
(expectancy and valence) affect goal acceptance, a Goal Theory variable; 
others have proposed the reverse. For instance, both Steers & Porter 
(1974) and Matsui et al. (1981) have asserted that goal attributes 
affect Expectancy Theory variables. 

Garland (1984) suggests that some of the conflicts may be resolved 

by a closer look at how the relation between effort-performance 

expectancy and performance has been treated in prior goal setting 

literature. He notes that 

researchers working in this area have failed to distinguish 
between expectancy-performance correlations within as 
compared to between goal groups. That is, although 
expectancy may be related negatively to performance across 
groups assigned goals of varying levels of difficulty, the 
relationship between expectancy and performance within goal 
groups may well be positive. Most researchers have simply 
ignored this distinction and collapsed goal groups when 
computing expectancy-performance correlations. This may 
account for the close to zero correlations that have been 
observed in previous studies. [Garland 1984, p. 80] 

The author suggests that higher expectancies should contribute to higher 
performance, provided all things, including goals , are equal. 

Implicit in Garland's (1984) arguments is the idea that for 
studies intending to measure expectancy-performance relationships, the 
varying of the level of goal difficulty cannot be considered a valid 


operationalization of the expectancy construct, even though increasing 
the level of goal difficulty lowers expectancy. According to him, such 
an operationalization would systematically report a false lack of 
expectancy-performance relationship. The precise manner in which 
expectancy fits into the goal setting phenomenon, as suggested by the 
integrating framework proposed in this study, shall be delineated in a 
forthcoming section of this report. 

Concerned with the controversy surrounding the expectancy- 
performance relationship, Janz (1982) manipulated subjective expectancy 
by providing differential feedback of performance scores to individuals 
assigned the task. His results provide partial support for the view 
that expectancy-performance relationship is inverted U-shaped. The 
findings also revealed that subjects with intermediate levels of 
expectancy outperformed those having low or high expectancies. 

Erez & Zidon (1984) proposed that goal acceptance could be a 
factor that moderates the relationship between expectancy and 
performance and could be a key factor in reconciling the inconsistencies 
between Locke's and Atkinson's models. They found support for their 
hypotheses that the expectancy-performance relationship is positive and 
linear for accepted goals; it is negative and linear for rejected goals; 
and the slope reversal (from positive to negative) is associated with 
transition from positive to negative values of goal acceptance. 
Earlier, Steers & Porter had expressed a similar idea in their 
conjecture that 


part of the seeming contradiction between Atkinson and Locke 
is that Locke was defining the right half of Atkinson's 
bell-shaped curve as either approaching or lying within a 
zone of impossibility, and thus it was not considered part 
of the goal difficulty-performance relationship. The 
remaining left half of Atkinson's curve closely resembles 
Locke's increasing linear function. [1974, p. 442] 

It requires to be mentioned at this juncture, however, that both Erez & 

Zidon (1984) and Steers & Porter (1974) are actually offering 

theoretical rationale for the inverted U-shaped curve in particular, 

rather than providing empirical support for the original Achievement 

Motivation Model. Atkinson's original model include proposals that 

generate U-shaped, inverted U-shaped, or flat (i.e., horizontal) 

expectancy-performance relationships. In other words, these studies 

have attempted to inquire beyond the boundary conditions of Goal Setting 

Theory and have proposed that the inverted U-shaped curve is a natural 

extension of Locke's model rather than provide theoretical insights that 

integrate Locke's conceptual foundations with those of Atkinson. 

Additionally, Atkinson's formulation specifies that performance 

and motivation should be maximum when subjective probability is 0.5, 

while Erez & Zidon (1984) found the threshold for goal rejection to be 

located between objective probabilities of 0.1 and 0.3. These authors 

did not report subjective probability-performance relationships, except 

to indicate that subjective probability correlated significantly with 

objective probability, with correlation coefficients very close to one. 

It must also be noted that, contrary to the observations of Erez & Zidon 

(1984), Locke and his associates have found no effect of goal acceptance 

on performance in several studies (Frost & Mahoney 1976; London & Oldham 


1976; Mento et al. 1980; Oldham 1975; Yukl & Latham 1978) . Finally, 
Erez & Zidon (1984) manipulated objective goal difficulty and did not 
report the correlations between subjective goal difficulty and 
performance. As Janz (1982) notes, it is the correlation between 
subjective probability and performance which is the focus of the 
controversy. Also, as Garland (1984) has observed, it is possible to 
obtain stable relationships between levels of goal difficulty and 
performance (i.e., objective probabilities of success and performance) 
across goal groups; while the expectancy-performance correlations may 
remain positive within goal groups. However, having stated all this, 
this study acknowledges the importance of the work of Erez & Zidon 
(1984) in contributing toward an integrated framework that will 
eventually reconcile a good measure of the conflicts examined thus far. 

Hollenbeck & Klein (1987) develop the idea proposed by Erez & 
Zidon (1984) further in distinguishing between goal commitment and goal 
acceptance. They adopt the definition of goal commitment originally 
suggested by Campion & Lord (1982) and indicate that "commitment implies 
the extension of effort, over time, toward the accomplishment of an 
original goal and emphasizes an unwillingness to abandon or lower the 
original goal" (Hollenbeck & Klein 1987, p. 212) . On the other hand, 
acceptance of difficult goals is supposed to refer merely to "the 
initial use of a goal assigned by another person as a referent. Goal 
acceptance does not necessarily imply that the individual is bound to 
the standard" (Hollenbeck & Klein 1987, p. 212) . These authors propose 
an Expectancy Theory Model of the goal commitment process in which, they 


suggest, goal commitment is a function of the multiplicative interaction 
of the "attractiveness of goal attainment" and the "expectancy of goal 
attainment." Goal commitment, in turn, moderates the relationship 
between goal level and task performance. The authors identify several 
antecedents of "attractiveness of goal attainment" and "expectancy of 
goal attainment," which they classify into situational and personal 
factors . 

One issue that this model has overlooked is the possibility that 
the goal levels themselves affect expectancies of goal attainment (with 
higher or more difficult goals contributing to lowered expectancies) . 
Although the authors acknowledge that feedback loops may exist, omission 
of this well known effect reduces the model to an unacceptable level of 
simplicity, so far as its capability to integrate conflicting paradigms 
is concerned. 



The self -efficacy construct is introduced within the context of 
goal setting since the conceptual framework in this study uses the 
construct to interrelate Goal Theory and Expectancy Theory variables. 
However, the treatment of the construct in this study will differ 
somewhat, especially with regard to the way it is measured. Details of 
these and other issues are described in this section. 

Self -efficacy is a key construct in Social Learning Theory 
(Bandura 1977a) . Bandura describes this theory as an unified framework 


for analyzing human thought and behavior. Social Learning Theory- 
emphasizes the importance of vicarious, symbolic, and self-regulatory 
processes in psychological functioning and adopts a perspective that 
attempts to explain behavior in terms of a continuous reciprocal 
interaction between cognitive, behavioral, and environmental 
determinants. In espousing the case for an interactionist position, 
Bandura (1977a, 1982b) suggests that the traditional view of 
interactionism, the idea that behavior results from the interaction of 
persons and situations, needs refinement. 

The interactionist approach is usually embodied in the relation 
B = / (P, E) , where B represents behavior, P the person, and E the 
environment. According to Bandura, the traditional interpretation of 
the interactionist position adopts a static perspective since it 
considers the person and the environment to be independent entities that 
combine to result in behavior. This notion of interaction, termed 
"unidirectional" by Bandura, is a severely limited view of behavioral 
phenomena, since person and environmental factors can and do influence 
each other. 

A more advanced but still somewhat deficient conceptualization is 
reflected in the "partly bi-directional" view of interaction. In this, 
persons and situations are acknowledged to be interdependent causes of 
behavior; but behavior is considered to be merely a consequence with no 
influence in the causal process [B =/ (P ■<""*■ E) ] . 


In the social learning view of interaction, known as "reciprocal 
determinism," behavior, personal factors, and environmental factors are 
considered to be mutually interactive determinants of each other. 

Bandura (1982b) suggests that it is difficult and perhaps 
impossible to study all aspects of this triadic reciprocality at the 
same time. Researchers may choose to focus on different segments of the 
interlocking reciprocality, without denying the existence and the 
possibility of influence by other parts of the chain. For instance, 
some investigators choose to inquire into the interaction between 
cognition and behavior and explore how conceptions, beliefs, perceptions 
of the self, and intentions affect behavior, on one hand; and how the 
consequences of behavior, in turn, alter thought patterns and affective 
responses. The author cites his own exposition of the self-efficacy 
construct (Bandura 1977b) as belonging to this domain. 

The Construct 

Self-efficacy is an individual-difference construct that refers to 
a person's perception of his or her own level of mastery within a 
limited task domain. In Bandura ' s language, "perceived self-ef f icacy is 
concerned with judgements of how well one can execute courses of action 
required to deal with prospective situations" (1982a, p. 122) . 

From the social learning perspective, knowledge and skills are 
necessary but not sufficient for successful performances or accomplished 
behavior. Evidence for this belief is found in the frequent suboptimal 
behavior of individuals, even when they possess full knowledge of the 


required behavior patterns. Bandura (1982a) argues that self-referent 
thought is a powerful motivator of the relationship between knowledge 
and behavior. Beliefs concerning a person's own capabilities are 
clearly very strong determinants of the inclination (or motivation) that 
individuals develop for either approaching and persisting in a course of 
action or avoiding a given behavior. Although self-ef f icacy does 
correlate with true ability to a considerable extent, this relationship 
is susceptible to a significant amount of influence from factors that 
relate to how the task is structured for the individual, and the manner 
in which it is learned or abstracted by the task performer. Bandura 
suggests that 

efficacy in dealing with one's environment, is not a fixed 
act or simply a matter of knowing what to do. Rather, it 
involves a generative capability in which component 
cognitive, social, and behavioral skills must be organized 
into integrated courses of action to serve innumerable 
purposes. . . . Self-efficacy judgements, whether accurate 
or faulty, influence choice of activities and environmental 
settings. People avoid activities that they believe exceed 
their coping capabilities, but they undertake and perform 
assuredly those that they judge themselves capable of 
managing. . . . Judgements of self-ef f icacy also determine 
how much^ effort people will expend and how long they will 
persist in the face of obstacles or aversive experiences 
[1982a, p. 122-123] 

The construct of self-efficacy has been found to be a powerful 
determinant of motivated performance in the realm of behavior 
modification within which it was developed. The original studies dealt 
with uncovering the antecedents of self-efficacy and measuring how well 
it correlated with performance, in settings involving modification of 
avoidance behavior of phobics (Bandura & Adams 1977; Bandura, Adams & 
Beyer 1977; Bandura, Adams, Hardy & Howells 1980; Lee 1984) . 


Subsequently, the effectiveness of the construct has been tested in a 
wide range of application domains including the prediction of writing 
performance (Meier, McCarthy & Schmeck 1984), behavior of sex offenders 
(Segal & Marshall 1986) , occupational preferences of college students 
(Betz & Hackett 1981; Clement 1987; Wheeler 1983), performance in 
athletic activities (Feltz 1982; Feltz & Mungo 1983; Feltz, Landers & 
Raeder 1979; Lee 1982; McAuley 1985; McAuley & Gill 1983; Weinberg, 
Gould & Jackson 1979; Weinberg, Yukelson & Jackson 1980), adoption of 
new technology (Hill, Smith & Mann 1987), and in the training for pain 
tolerance (Manning & Wright 1983; Vallis & Bucher 1986), as well as in 
the context of training for weight control/reduction (Bernier & Avard 
1986; Glynn & Ruderman 1986), controlling the smoking habit (Condiotte & 
Lichtenstein 1981; DiClemente 1981; DiClemente, Prochaska & Gibertini 
1985; Godding & Glasgow 1985)), and treatment for alcoholism (Rollnick & 
Heather 1982) . 

Within the context of organizational behavior, self-efficacy has 
been found to be a determinant of goal-choice (i.e., the level of self- 
set goal) in a performance task (Locke, Frederick, Lee & Bobko 1984) . 
This study indicated that self-efficacy continued to be a significant 
predictor of future performance even when past performance was 
controlled. Based on their observations, these authors speculated that 
the self-efficacy concept could provide an integrating mechanism between 
the goal-setting and social-learning-theory approaches to task 
performance. Bandura & Cervone observed that "self-evaluative and self- 
efficacy mechanisms mediate the effects of goal systems on performance 


motivation" (1983, p. 1017). Locke, Lathams Erez (1988) consider both 
self-efficacy and expectancy of task success to be antecedents of goal 
commitment, which in turn is a critical requisite for the goal setting 
effect to occur. Gist (1987) argues that the concept of self-efficacy 
is one of enormous significance to the domain of organizational behavior 
and human resource management and that it should receive far more 
attention than it has received thus far. 

In view of the diversity of the application areas of the self- 
efficacy concept and the robustness of its theoretical foundations, it 
appears to hold substantial promise with regard to its potential for 
contributing toward an overarching framework. Even before delving into 
the intricacies regarding definitional and measurement issues, it 
appears to be intuitive that the results of studies reviewed earlier 
indicating the salutary effect of high expectancies on performance are 
not in contradiction with studies reporting enhanced performance on 
account of high self-efficacies. Similarly, those studies that have 
hypothesized or reported high self-esteem to be an antecedent of 
improved expectancy and performance could also be interpreted as 
providing complementary evidence for the self-efficacy effect. However, 
it is not as if that the interrelationships between these constructs are 
completely clear from their prior treatment in the literature. In 
particular, Self-Efficacy Theory was often positioned as a rival 
paradigm to Expectancy Theory (e.g., in Lee 1984). However, a 
substantial amount of the conflicts (or consonance) depends on the exact 
manner in which these constructs are interpreted. These issues, 


specifically comparisons of self-efficacy with related constructs and 
the corresponding theoretical and measurement concerns, are explored in 
the following sub-sections. 

Self-Efficacy and Expectancy — Theoretical Issues 

Both self-efficacy and expectancy are constructs that invoke 
cognitions concerning anticipated performance. It is not surprising, 
therefore, that there should exist a close correspondence between them. 
In fact, it should be possible to delineate the relationships between 
all such concepts involving self-related cognitions (e.g, self-esteem) 
which have some bearing on behavior. 

One problem that arises in the attempt to compare and contrast 
such constructs is the fact that often their treatments in the hands of 
different researchers have been dissimilar. Also, some of them have 
undergone evolutionary changes ranging from the subtle to the 
substantial during the course of their existence. As a result, it 
becomes almost impossible and, to some extent, pointless to venture into 
a comparative study unless the concepts are formally and strictly 

As mentioned before, prior to the refinement proposed by Porter & 
Lawler (1968) , expectancy was regarded as the subjective probability 
that effort on a task would lead to a given outcome . For this reason, 
Bandura (1977a & b) refers to this traditional view of expectancy as 
"outcome expectancy." He argues that self-efficacy, also referred to as 
"efficacy expectations," is a better predictor of behavior than outcome 


expectancy. According to him, individuals may believe that a particular 
course of action will result in certain outcomes and yet doubt whether 
they themselves can execute those actions. He represents the 
differentiation between outcome expectancy and self-efficacy as the 
difference in the links of the causal path: Person - Behavior - Outcome. 
Efficacy expectations refer to the subjective contingency between Person 
and Behavior; while outcome expectancy refers to the subjective 
contingency between Behavior and Outcome . 1 

In this respect, self-efficacy bears a striking resemblance to the 
concept of expectancy I or E-I proposed by Porter & Lawler (1968) and 
elaborated by Campbell et al . (1970). Since expectancy I refers to the 
subjective probability that effort will lead to a specific level of 
performance , the similarity is quite obvious. However, researchers who 
have noticed this resemblance also emphasize that fundamentally self- 
efficacy is a more broad-based and inclusive concept than expectancy I 
(Gist 1987; Locke et al . 1984). For instance, Gist indicates that the 
definition of self-efficacy "implies that judgements of efficacy depend 
on more than effort considerations and, thereby, subsume variables not 
included in El" (1987, p. 477) . Also, in most studies, self-efficacy 

Lee (1984) attempts to demonstrate empirically that efficacy 
expectations are superior predictors of performance than outcome 
expectancy. However, other researchers (Eastman & Marzilllier 1984; 
Marzilllier & Eastman 1984) have criticized these ideas on grounds that 

(1) a task cannot be defined without some reference to outcome, and 

(2) outcome expectancy does contribute to future behavior irrespective 
of whether individuals believe in their competence to perform the task. 
(See also Bandura's [1984] response to these criticisms). 


measures include expectation ratings for a wide range of performance 
levels; while each expectancy measure corresponds to only one specific 
performance level. 

For the purpose of this study, self-efficacy is regarded as the 
individual's perceived sense of mastery in a given performance domain, 
in accordance with Bandura ' s (1982a) definition. Expectancy, or 
expectancy of task success, specifically refers to an individual's 
subjective probability that he or she will achieve a specific level of 
performance on a given task. Thus, as the characteristics of the task 
are varied, expectancy of success at the task may also be expected to 
change. Self-ef f icacy, on the other hand, is conceptualized as a more 
global and stable variable whose values remain relatively unaltered over 
a range of performance levels in a given task domain. 

The global aspect of self-efficacy is embodied in the idea that an 
individual's perceived sense of mastery in a given performance domain 
may have generalization effects on similar or related domains of 
performance. For instance, an athlete who considers himself or herself 
competent in a particular sporting activity is likely to transfer (or 
generalize) some of that sense of competence to related or similar 
sporting activities. This generalization effect would nevertheless be 
dependent upon how similar the performance domains (across which the 
transfer effect occurs) are perceived to be; or how performance domains 
are demarcated. The more similar the different performance domains are 
perceived, the greater will be the generalization effect. Also, if two 
activities with highly similar physical characteristics are classified 

as belonging to separate performance domains, then the generalization 
effect across those domains will also be large. 

Self -efficacy is also a relatively enduring variable. Individuals 
develop impressions of personal efficacies over a period of time through 
several sources of information, included among which are enactive 
mastery (i.e., sense of accomplishments derived from actual successful 
participation in an activity) , vicarious experience, verbal persuasion, 
and emotional arousal (Bandura 1977a & b, 1982b; Bandura, Adams & Beyer 
1977; Gist 1987) . Impressions of efficacy are, therefore, composite 
indices of individuals' sense of competence. In contrast, expectancy of 
task success is directly related to how difficult the task (and the 
assigned level of performance or goal) is perceived to be. Manipulating 
expectancy is possible by varying the characteristics of the task. 
Self-efficacy, on the other hand, may be altered only by changing the 
individual's overall impression of his or her general capabilities in 
the performance domain in question. The idea that self-ef f icacy is 
relatively enduring should not however be construed to imply that it is 
impervious to change. Also, unlike personality variables, it is not an 
inherent property of the individual that has little or no susceptibility 
to external influences. As Bandura emphasizes, "self-referent thought 
is indexed in terms of particularized self -percepts of efficacy that can 
vary across activities and situational circumstances rather than as a 
global disposition assayed by an omnibus test" (1982, p. 124) . 


Self-Efficacy and Expectancy— Measurement Issues 

Traditionally, expectancy measures have been obtained by having 
subjects report probability estimates of their attaining an outcome (or 
the achieving of a performance level, in the case of E-I) in the context 
of a task. For instance, if a respondent reported that he or she 
estimated his or her chances of attaining a given performance level was 
70%, then expectancy (E-I) was considered to be 0.7. This method of 
measuring expectancy is consistent with the way in which the construct 
has been defined in this study. 

In the case of the measurement of self-efficacy, the issue is not 
as simple. Efficacy has also been measured using probability estimates 
or composites of probability estimates (Bandura 1980; Bandura & Adams 
1977; Bandura & Cervone 1983; Bandura & Schunk 1981; Bandura, Adams & 
Beyer 1977) ; as well as mean responses on specially constructed scales 
(Barling & Beattie 1983; DiClemente, Prochaska & Gibertini 1985; Glynn & 
Ruderman 1986; McAuley & Gill 1983) . It is my contention that the 
typical measures of self-efficacy utilized by Bandura and his associates 
capture only a narrow aspect of the self-efficacy construct. 

Consider, for example, a study by Bandura & Cervone (1983) in 
which self-efficacy was measured by having the subjects rate how certain 
they were about attaining each of the several goal levels on 100-point 
scales. That is, although each of the respondents were assigned only 
one particular goal level, their estimates of the probabilities of goal 
attainments were obtained for several other performance levels which 
were not assigned to them. This measure of self-efficacy is no 


different than a measure of expectancy with the exception that self- 
efficacy was measured with respect to a range of performance levels in 
addition to the assigned goal level. In other words, self-efficacy is 
measured merely as a composite of expectancy scores or sometimes as a 
mean of all the expectancy estimates measured. 

These composite expectancy scores (or means of scores) are likely 
to be higher for those high in self-efficacy (on account of the reasons 
mentioned as follows) and therefore their use as surrogates for self- 
efficacy should not have caused inconsistent results. Goal theory 
predictions would suggest that expectancies of task success diminish as 
goals are made more difficult. Therefore, measures of expectancies for 
a series of goals of increasing difficulty should generate downward 
sloping curves (for all individuals) . However, individuals with high 
self-efficacies are likely to have downward sloping curves that have 
larger intercepts and, perhaps, less negative slopes (i.e., less steep 
slopes) in comparison with those low in self-efficacy. 

In spite of the fact that there is probably no real danger of 
obtaining inconsistent results on account of measuring self-efficacy as 
described, the question is one of construct validity. It has been 
reiterated on several occasions that the self-efficacy construct 
subsumes many more facets of the behavioral phenomena than expectancy 
does (Gist 1987) . Bandura emphasizes that the focal interest of Self- 
Efficacy Theory is "the dynamic interplay among self-referent thought, 
action, and affect" (1982a, p. 124) . This construct is thought to 
account for phenomena as varied as coping behavior produced by different 


modes of influence, level of psychological stress reactions, self- 
regulation of refractory behavior, resignation and despondendency to 
failure experiences, among others (Bandura 1982a) . The measurement of a 
construct as broad-based as this ought not to be limited merely to 
expectancy ratings. 

I contend that measures of self-efficacy should incorporate 
specially constructed scales that tap into the individual's composite 
impression of his or her capability in the task domain in question. The 
conceptual model in the following chapter espouses the idea that 
nomologically self-efficacy may be an antecedent of composite or mean 
expectancy scores (and proposes exactly that in a later section of this 
report); and, therefore, expectancy measures may, if required, be used 
as manipulation checks for self-efficacy (as has been done in Weinberg, 
Gould & Jackson 1979) . 

Bandura acknowledges that "self-efficacy scales vary in their 
structure depending upon the domain of functioning and the specificity 
with which it is being examined" (1984, p. 241) . Several researchers 
have consequently developed scales specific to their area of inquiry and 
some have even tested the validity of the scales constructed 
specifically to assess self-efficacy within a given performance domain 
(e.g., Glynn & Ruderman 1986). 

In sum, then, self-efficacy is a construct that is conceptually 
distinct from expectancy. It also subsumes the expectancy variable and, 
therefore, previous attempts to measure self-efficacy by expectancy 


scores over a performance range have generated results not inconsistent 
with theory. 

Self -Efficacy and Self-Esteem 

There exist studies in the area of salespeople's motivation which 
indicate that self-esteem is often a valid antecedent of expectancy. 
The interrelationship between these two constructs and self-efficacy may 
be viewed as a hierarchical progression of antecedents, in which self- 
esteem is the most broad-based which subsumes both self-efficacy and 
expectancy; while self-efficacy is the second in line and is an 
antecedent of expectancy. Schematically, this can be represented as: 
Self-esteem > Self-ef f icacy > Expectancy. 

This schematic representation only suggests the ordering in terms 
of the direction of influence so far as these three variables are 
concerned. It is not implied in any way that this represents a complete 
picture with regard to all possible antecedents that either expectancy 
or self -efficacy can have. The idea expressed in this representation is 
that self-esteem is the broadest of the three variables in question; and 
it is imaginable that an individual with high self-esteem may have an 
elevated sense of efficacy compared with an individual low in self- 
esteem. In other words, it is proposed that previous studies that have 
demonstrated positive effects of self-esteem on expectancy are not only 
consistent with this scheme but can also be interpreted in its light. 


This chapter describes a conceptual model that integrates previous 
research on the overall goal setting phenomenon. The model identifies 
relevant motivational mechanisms that intervene in the effect of an 
externally set goal on performance. By providing an overarching 
theoretical framework, the model accounts for the putatively conflicting 
viewpoints of different research traditions that were examined in the 
previous chapter. Additionally, it generates theoretical propositions 
concerning different effects of goal setting. This chapter describes 
the conceptual model and develops the set of propositions. In the next 
chapter, the methodology to be employed for testing a select group of 
these research propositions is discussed. 

The following section of this chapter briefly reviews some of the 
earlier attempts to account for or explain the goal setting effect. In 
the process, the necessity of developing a more complete, conceptually 
oriented framework is highlighted. 

For the purpose of clarity and also because of the opportunity it 
provides for detailed examination of the several aspects involved in 
goal setting, the conceptual model will be described in a number of 
progressive stages. In a subsequent section, this scheme is described 



more fully and its advantages are outlined. Finally, the different 
aspects of the model are discussed and integrated. 

Object ivfis o f the Model 

As mentioned in Chapter 2, despite the proliferation of studies 
that have attempted to demonstrate the enhanced effects of difficult and 
specific goals on performance, theoretical inquiries into the mechanism 
responsible for the phenomenon have been rare. 

Locke's earlier studies (1966, 1968) suggested that difficult 
goals are responsible for generating and elevating behavioral 
intentions. Intention to perform, in turn, contributes to actual 
performance. Although, there has been some discussion concerning 
identification and acceptance of goals in terms of cognitions and 
evaluations, the goal-intention-performance link was viewed as adequate 
"explanation" for the Goal Theory propositions. 

For example, Locke & Latham (1984) refer to goal setting as a 
technique. In their opinion, the following "causes" or reasons are 
responsible for the fact that the technique produces its intended 
results: 1) specific goals direct action more reliably than vague or 
general goals, 2) goal specificity results in clear expectations, and 
3) the harder or more challenging the goal, the better the performance. 
The preceding rationale is somewhat tautological since there exists an 
obvious confound between what is being explained and that which is being 
used to explain it. 

However, Locke et al . (1981) do mention that they view goal 
setting as primarily a motivational mechanism that necessarily involves 


cognitive elements as well. They suggest that "the goal setting 
mechanisms" involve the following: 1) goals direct attention, 2) goals 
mobilize effort, 3) goals induce persistence in effort, and 4) goals 
motivate strategy development. 

The studies previously reviewed and cited in this chapter do not 
follow a systematic method in providing a conceptual explanation of the 
goal setting phenomenon, because researchers in this tradition have 
focussed on the identification of the correlates and the epiphenomena 
associated with the goal setting effect rather than on conceptual 
explanation. Consequently, in their attempts at describing the effect, 
these researchers have amassed a large number of variables associated 
with the goal setting effect. However, attempts at orderly structuring 
of the theoretical explanation of the phenomenon fall far short of their 
impressive record of empirical studies. 

One of the objectives of this conceptual framework is to examine 
and account for the goal setting effect and related phenomena in terms 
of the intervening cognitive and motivational processes in a systematic 
manner. While the previous research has focussed mainly on describing 
haw the goal setting effect works, this model predominantly inquires 
into why the effects occur as they do, and the conditions under which 
they are likely to change. 

.Scheme Adop ted for Describing the Conceptual Model 

As a brief and somewhat simplified introduction, the integrated 
model suggests that externally set goals affect performance through 
several intervening variables and multiple pathways. The ultimate 


impact on performance is contingent upon the interactive effects of 
factors that moderate the relative strengths and direction of the 
pathways . 

As can be expected, externally set goals affect performance in 
multiple ways and combine with other factors in the process. My 
exposition of the goal setting effect, however, will initially consider 
only the "primary effects" in isolation and later explore the relevant 
interactive effects. In other words, I will begin to examine why 
increasing externally set goals improves performance, without consi- 
derations of any other situational factors, such as rewards, task 
structure, etc., or any inter-individual differences. 

The development of the framework begins with a reexamination of 
the goal setting effect — the phenomenon of more difficult goals 
contributing to better performance. The rest of the model is built 
around this central "pathway" or route. 

The following section contains a detailed description of the 
mechanisms involved in this central link, referred to as the Extended 
Compliance Path. 

The Ext ended Compliance Path 

Overview of the Path 

This pathway (as one of the several through which such effects may 
occur) suggests that the level of externally assigned performance target 
corresponds to a cognitive encoding of the external goal by the 
individual to whom the task is assigned. This internal representation 
of the goal influences his/her intention to perform at the assigned 


task. The effect on intention to perform is carried over to the actual 
resources expended by the individual and that, in turn, contributes to 
performance. This extended compliance pathway is schematically 
represented in Figure 3.1. 

Set Goal 



of Goal 

Intention to 



Figure 3.1: The Extended Compliance Path 

Cognitive Rep resentation of Gnxl 

Externally assigned goals are necessarily exogenous variables in 
so far as they are manipulated and varied in an environment external to 
the individual's cognitive system. Also, the level of the goals can be 
quantitatively measured (in terms of physical units) without any 
reference to the internal system. It is proposed that when a task is 
presented to a task performer, the individual in question forms a 
composite impression of the task from all the information regarding the 
task that is available to him or her. One very important piece of 
information that is utilized by the individual in forming a cognitive 
representation of the task is the actual level of the assigned goal. 
The internal representation of the task is, in a sense, a cognitive 
mapping of the actual task presented to the individual. 

Similarly, the actual level of the goal assigned can be viewed as 
influencing an internal representation of the task goal . It can also be 

thought of as being "assigned" a certain value along a (cognitive) 
dimension of the task which represents goal difficulty or level of goal. 
It is not only unnecessary for this study but also beyond its scope to 
inquire into the detailed architecture of such a cognitive structure. 
It suffices to mention, however, that if external goals influence 
intentions, then they must also be "encoded" cognitively at some level 
before they can influence intentions or performance. 1 

This encoded or internal representation of goal can be thought of 
as an individual's understanding of what the goal level stands for at a 
general IrvpI . That is, this primary encoded representation takes place 
even before the person's own ability concerns are considered and is only 
a general mapping of the external goal in the abstract. Individuals may 
be visualized as being able to appreciate the implications of different 
levels of goals even for a task that they will never personally 
confront. The abstract appreciation that increasing goal levels at any 
task generally corresponds with increased effort requirements on part of 
the task performer is a manifestation of internal representations of 
goals at the global or general level. This is not to say that people 
ordinarily do not factor in ability concerns for tasks that they will or 
will not confront— they probably always do. But the mechanism being 
proposed here suggests that evaluation of the task or goal difficulty in 

The scheme adopted here pursues an "encoding-processing-output" 
orientation toward analyzing information processing issues. This is not 
only conducive toward analyzing motivational mechanisms that have 
behavioral manifestations but is also consistent with a large body of 
literature that adopts this model. 

5 9 

relation to self-appraised ability occurs subsequent to a relatively 

primary encoding of the goal. In other words, an individual's response 

to a task goal is contingent upon how the goal itself is encoded at a 

general level as well as other factors including cognitions about the 

self. It is theoretically important to study each of these separate 

effects in isolation (as different links of the entire model) in order 

to obtain a clear picture. 

Proposition 1: The level of performance goal assigned to an 

individual in the context of a task is encoded 
internally, such that the actual or physical level of 
the goal (measured in quantitative terms) corresponds 
to the internal representation of goal. 

The preceding idea is schematically expressed in Figure 3 . 1 by the 

link between the "level of externally set goal" and "internal 

representation of goal." 

Effect on Intention— The Compliance Mechanism 

One of the ideas expressed explicitly by Locke (1966, 1968) is 
that individuals respond to external goals by raising their intentions 
to perform as long as the goals are accepted. He indicates that 
difficult goals hold more "challenge" for the participants but does not 
extend or develop this concept any further. The notion of challenge 
associated with difficult goals involves considerations of intrinsic 
rewards. Such matters will be addressed and integrated with the model 
in a later section. it requires to be mentioned, however, that apart 
from the effects of "challenge" there may exist other mechanisms through 
which the primary goal setting effect can take place. 


It is significant that the response of subjects to performance 
targets even in very simple and unexciting tasks have consistently 
supported the Goal Theory postulates in several studies. The earliest 
and some of the simplest applications of goal setting involve 
experiments in which subjects were asked to perform at a certain level 
in a simple motor task (e.g, Eason 1963; Eason & White; 1961) or a paper 
and pencil task involving simple arithmetic (e.g., Mace 1935) or a task 
such as listing of words (Locke 1966). The subjects' actual 
performances were found to correspond closely with the assigned goals. 

It is difficult at times, if not impossible, to extract any sense 
of how a difference between two goal levels in a very dreary task could 
generate so much "challenge" that it could account for the entire 
effect. It is the existence of the goal setting effect in these tasks, 
however, that lend an indication or clue regarding one mechanism by 
which goals affect performance. 

An important aspect of social behavior that is of interest in this 
context is that the subjects in these studies seem to respond 
"adaptively" to whatever the task demands are (within a reasonable 
limit) and try to comply with that which is asked of them. This idea is 
also evident in Locke's concern that goal setting results may appear 
"obvious" in some cases since "the subjects were presumably doing what 
they were told" (1966, p. 62) . It is proposed here that (in addition to 
the other pathways through which goals may affect performance) some of 
the results seen in goal setting studies owe much to what can be termed 
as the "compliance effect." This effect describes the phenomenon in 
which individuals redirect their intentions and efforts and modify their 


intended responses to match the changing requirements of the task 

(within certain bounds) . 

Although by no means a depiction of the entire range of motivated 
behavior, this idea finds support in the work of Feldman and Lynch 

(1988) who propose that behavior is at least partially "mindless" in the 

sense that much day-to-day action occurs without deliberate intent and 

without awareness of controlling factors. As an example, these authors 

suggest that it is unlikely that people choose to attend work every day 

after consciously evaluating the advantages of doing so. They respond 

without deliberating at any length to what they perceive as certain 

environmental demands as they become aware of them. 1 

Proposition 2: The individual's internal representation of goal has a 

direct positive effect on the his/her intention to 
perform; individuals modify their behavior simply in 
response to the task requirements — the goal level, in 
this case. 

The details regarding the conditions under which the preceding statement 

will (or will not) be valid shall be amplified later. The preceding 

proposition embodies the compliance mechanism and is depicted 

schematically in Figure 3 . 1 by the link between "internal representation 

of goal" and the "intention to perform." 

1 This does not imply, however, that this behavior is "irrational" 
or unusual in any other way. Exploration of the origins of such 
behavioral habits can reveal their underlying causes or reasons. At a 
molar level, however, the effect is analogous to pure compliance with an 
external task requirement. 


Effect Of Int e ntion nn Psrfnrm^ nrp 

The effect of intention on performance is relatively 
straightforward and is analogous to the effect of behavioral intention 
on behavior. In our model, however, "intention to perform" and actual 
performance is spanned by another behavioral indicator— resources 
expended (see Figure 3.1). Resources include the effort that is 
expended by the individual. This effort is the behavioral manifestation 
of "intention to perf orm"--the composite cognitive factor that 
summarizes the multiple effects of goals. In addition to actual effort, 
resources may include cognitive activities such as planning. Although 
this thesis focuses more attention on the effort expended (as a 
consequence of intention and antecedent of performance), inclusion of 
more strategic aspects of output performance in the model lends it a 
higher degree of generality. Besides, behavioral manifestations may be 
thought of as consisting of both magnitude and direction components. 
Finally, a recent goal setting study (Early, Wojnaroski & Prest 1987) 
has identified strategy development as an intervening variable that may 
aid in our understanding of the relevant processes. 

The rationale for introducing "resources expended" as an 
additional intervening variable is that the relationship between 
intention and performance is also subject to certain moderating 
influences (to be discussed later) . 

Proposition 3: Intention to perform in the context of a task directly 

influences the resources (effort and strategic 
planning) expended by the individual. 

Proposition 4: Resources actually expended by the individual in the 

context of a task contributes directly to the level of 
performance achieved in the task. 


The preceding propositions are represented in Figure 3.1 by the links 
between "Intention to perform," "Actual resources expended," and 
"Performance . " 

Figure 3.1 is a schematic summary of all of the four preceding 
propositions and represents the Extended Compliance Path in its 

Moderating Influences on the Extended Compliance Path 

As has been indicated earlier, the level of externally set goal 
affects performance through multiple and interactive mechanisms. 
Accordingly, we may expect that there will exist a number of moderating 
influence on the Extended Compliance Path described in the preceding 
section and represented in Figure 3.1. 

Included among the important moderators of the Extended Compliance 
Pathway are Goal Specificity, Perceived Extrinsic Rewards, and 
Expectancy. In this section, the influences of these moderators on the 
central pathway are elaborated by specifying how these moderators 
interact with and modify the Extended Compliance Mechanism. Two of the 
three aforementioned moderators, viz., "extrinsic rewards" and 
"expectancy, " may themselves be contingent upon the externally set goal 
under special circumstances. Following sections of this chapter address 
the Extrinsic Reward Pathway and the Goal-Expectancy Pathway in their 
entireties . 


Effect of Goa l Specificity 

Goal Theory propositions suggest that specific goals are more 
conducive to generating superior performance in comparison with vague 
goals. Locke and his associates refer to goals that are described in 
concrete, quantitative terms as specific goals; while performance goals 
conveyed to task performers by such advice as "do your best" have been 
referred to as vague goals (Locke 1968; Locke et al . 1981). Locke's 
postulates suggest that goals that are both specific and difficult at 
the same time are more effective than easy or vague goals with regard to 
the level of performance generated. However, these vague goals (or "do- 
your-best goals") may often result in higher performance than what can 
be achieved by moderate or easy externally set goals (Locke, Mento & 
Katcher 1978) . 

The effect of "do-your-best" goals relative to specific goals can 
be best understood in terms of the compliance mechanism and the internal 
representation of goal. The compliance mechanism suggests that 
individuals tend to adapt their intended efforts to match their 
perception of the task goals. It is proposed that recommendations to 
task performers to "do their best" convey to them the impression that 
they have some amount of freedom in choosing or shaping the task goal. 
It is more than likely that the subjects pick up signals regarding the 
importance of the task in general and the task goal in particular and 
respond to the overall message rather than interpret the assignment in 
literal terms. In so far as the semantics and connotations of the 
phrase "do your best" diverge from its implications when examined word 
by word, a goal associated with such a description lacks specificity. 


The preceding argument reiterates the idea that the internal 
representation of goal difficulty is contingent upon the information 
available with regard to the task goal. If the cues associated with the 
task goal contain conflicting information or unintended messages, then 
the task performers' encoding of the task will be affected 
correspondingly. To the person who assigns the task, however, it will 
appear to be an anomalous case. 

Locke's treatment of goal specificity concerned only the 
difference between difficult, quantitatively-stated performance goals 
and "do-your-best" goals. As we can see, the notion of specificity (or 
clarity) may be of importance even beyond this particular issue. For 
instance, it is a matter of conjecture as to which of the two goal 
conditions between a very easy, specific goal and the assignment 
"perform at your minimum level" would actually result in the lower 
performance. It is likely that the task performers will interpret the 
phrase "perform at your minimum level" in a similar fashion. They could 
decipher it as an indication for them to work at a relaxed pace but 
within some acceptable limits (that they may set for themselves or 
perceive others to have set for them) . It may be hypothesized that in 
such a situation the task performers will not perform at zero level (the 
physical minimum possible) which is what the task goal literally 
connotes . 

A practical implication of this observation is that even when goal 
levels are communicated to subjects in quantitative terms but are in 
complex forms (e.g, as in such descriptions as "within one standard 
deviation of the mean of previous trials" or "at the 80 percentile level 


or higher") or in broad terms (such as "significant improvement over 
previous attempts"), it will contribute to an attenuation of the goal 
setting effect and result in a greater variability of the internal 
representation of goal among the respondents. 

The advantage of describing the task goal in specific terms (over 
describing it in vague or general terms) roughly corresponds to the 
superior effect of assigning a task in conjunction with a performance 
goal relative to assigning it without explicit performance targets. In 
the first case, the task goal contains more descriptive information that 
aids in the formation of a more coherent internal representation. In 
the second case, a more complete cognitive mapping of the task is 
possible. This is also in accordance with another related aspect of 
task goals elaborated in the following paragraph. 

Task difficulty is believed to influence the amount of attention 
allocated by the individual (Kahneman 1973) . In so far as the tasks 
described in terms of performance goals are allocated more attention by 
the task performers (and as long as difficult goals tend to command more 
attention) , task goals or performance targets may be thought of as 
possessing attention enhancing properties. Enhanced attention, in turn, 
aids in the individuals' development of the intention to perform and 
consequent performance. 

It stands to reason then, goal specificity may interact with the 

effect of externally set goal as suggested by the following proposition. 

Proposition 5: Goal specificity (or clarity of assigned goal) 

moderates the relationship between externally set goal 
level and the internal representation of goal as 
follows. The effect of externally set goal on its 


internal representation is stronger when goal 
specificity is high than when goal specificity is low. 

Figure 3.2 depicts the interaction schematically. 

Set Goal 




of Goal 

Intention to 




Figure 3.2: The Moderating Effect of Goal Specificity 

Effect of Extrinsic Rewards 

For the purpose of this model, rewards are defined as consequences 
of task accomplishment that are evaluated positively (and, consequently, 
desired) by the task performer. Extrinsic rewards are those which are 
administered and controlled by an agent external to the individual and 
are contingent in some manner upon the evaluation of the performance by 
that (or another external) agent. 

Intrinsic rewards, on the other hand, refer to the satisfaction or 
enjoyment derived from actual participation in the task itself. They 
are predominantly contingent upon the interaction of the task 
characteristics and the characteristics of the task performer rather 
than an external agent. 1 

The supervisor of a task (as an external agent) does possess some 
power to alter task characteristics which may indirectly influence 


Tasks that have absolutely no extrinsic or intrinsic rewards 
associated with them are not only rare but can exist only as theoretical 
abstractions. Usually, some form of extrinsic reward, however subtle, 
accompanies almost all task assignments. These rewards may not be 
tangible but will still qualify as extrinsic rewards as long as the 
recipient is dependent upon another person's evaluation and discretion 
in order to obtain it. Examples of intangible extrinsic rewards may 
include social approval, admiration, praise, respect, etc. 

By definition, extrinsic rewards are those that are in some way 
tied to an evaluation of the task performance by an external agent, 
usually the same person or agent that assigns the task. For tasks that 
are assigned in terms of performance targets, attempts at realizing 
external rewards involves responding actively to the assigned goal 
level. Consequently, extrinsic rewards are likely to boost the effect 
of the compliance mechanism. The following proposition expresses this 
idea more formally. 

Proposition 6: Perceived extrinsic rewards moderate the compliance 
mechanism as follows: high levels of perceived 
extrinsic rewards result in stronger positive effects 
of internal representation of goal on intention to 
perform than low levels of extrinsic rewards. 

Figure 3.3 is a schematic representation of the moderating effect of 

perceived extrinsic rewards on the compliance mechanism. 

intrinsic rewards. Their administration and control, however, are 
unlikely to be related to evaluation of performance. 



Set Goal 



of Goal 


Intention to 



Figure 3.3: The Moderating Effect of Extrinsic Rewards 

Figure 3.4 depicts a hypothetical interaction of Internal goal 
representation and Perceived extrinsic rewards on Intention to perform. 

Intention to 




Extrinsic Rewards 

Internal representation 
of goal difficulty 

Figure 3.4: An Example of a Hypothetical Interaction Between Internal 
Goal Representation and Perceived Extrinsic Rewards 

An important feature of the preceding interaction is that when 
external rewards are the only source of rewards, at the same time, the 
level of perceived external rewards are held at zero, the compliance 
effect will vanish. That is, in the absence of any reward whatsoever, 


the motivation to comply in response to increasing goals will be 
nonexistent . 1 

Similarly, when the externally assigned goal is set at zero, 
increased perceived extrinsic rewards should not raise intention to 
perform. The preceding scenario corresponds to a situation in which 
(increasing) extrinsic rewards are perceived to be conditional upon a 
zero level of target performance; i.e., the extrinsic rewards are 
perceived to be free. Since, no performance is requested of the 
individual in exchange for the external rewards, no intention (for 
performing a task) will be evoked. 

The Extended Compliance Mechanism described earlier and depicted 
in Figure 3.1 (and embodied in propositions 1 through 4) implicitly 
assumes a fixed, finite level of perceived extrinsic rewards. 

Percep tion of extrinsic rewards . It requires to be emphasized at 
this point that it is the perception of extrinsic reward (rather than 
the actual reward) that matters so far as the preceding moderating 
influences are concerned. Undoubtedly, the actual external reward 
level — an operational variable — strongly influences how it will be 
perceived. Nevertheless, factors such as individual characteristics and 
message factors (involved in communicating and administering the reward) 
are likely to interact with (i.e., moderate) this relationship. 

1 For the purpose of making a point, it is assumed that in this 
case there exists no intrinsic rewards either. Also, the threats of 
punishments (which may be considered to be analogous to negative 
rewards) for noncompliance are ignored. 


For instance, those high in need for achievement (n-Achievement) 
or need for affiliation (n-Af filiation) are likely to perceive 
intangible extrinsic rewards (such as praise, admiration, etc.) to be 
more positively valenced than those who are low in these needs. The 
rationale for this argument is that the intangible rewards usually cater 
to the needs associated with the self-system: the higher the need in the 
self -system, the higher will be its value to the individual. (The 
economic analogy of higher demand contributing to higher price is 
evident) . 

Self-esteem is likely to moderate the perception of intangible 

(self-related) extrinsic rewards for similar reasons. Those high in 

self-esteem are likely to have less use or inclination for externally 

administered positive evaluations of their self -systems . 

Proposition 7: Both individual and reward characteristics moderate 

how external rewards will be perceived by the task 
performer (to whom the reward is administered) . 

Proposition 7a: The individual's perception (or evaluation) of the 

externally administered rewards will be moderated by 
certain individual characteristics as follows: 
(i) intangible extrinsic rewards will tend to be more 
positively evaluated by those who are high in 
n-Achievement and n-Af filiation than those who are low 
in these needs; 

(ii) those high in self-esteem will evaluate 
intangible extrinsic rewards less positively than 
those low in self-esteem. 

Proposition 7b: Individual differences (such as differential 

preferences for different kinds of external tangible 
rewards) will moderate the perception of external 
tangible rewards. 

Proposition 7c: Message characteristics (that highlight or underplay 

the importance of different features of extrinsic 
rewards) will also moderate the relationship between 
the externally administered reward (both tangible and 
intangible) and its perception. 

External Rewards 


Message and 




Extrinsic Rewards 

Figure 3.5: Moderating Influences on Perceived Extrinsic Rewards 

Effect of Expectancy 

Expectancy, the subjective probability of success at a task, has 
been found to exert a strong impact on individuals' motivation for 
behavior under a wide array of circumstances. According to Garland 
(1984) , expectancy models predict that, all things being equal, 
expectancy will positively influence performance. That is, if two task 
situations are identical in all respects except for the individual's 
expectancy estimate, then the individual will be motivated to perform at 
a higher level for the task associated with the higher expectancy. It 
is proposed that the compliance mechanism (which suggests that task 
performers tend to match their intended performance levels to their 
perception of the externally assigned goal) will also be moderated by 
expectancy in a similar fashion. If the individual perceives his or her 
chances of succeeding at a task (i.e., achieving the assigned goal) to 
be high then it may be argued that he/she will be more strongly 
motivated to comply than when the subjective expectancy estimates are 
low. In fact, if the expectancy of task success is extremely low (i.e., 


for task goals that are perceived to be near impossible to accomplish) , 

increasing the level of goal difficulty may detract from the individuals 

motivation to comply. 

The preceding argument embodies the assumption that individuals do 

not comply "mindlessly" for all tasks or all levels of difficulty at 

routine tasks. They may be expected to comply with assignments only for 

those tasks and goal levels with which they are familiar and in which 

they perceive to have acceptably high levels of expectancies. This 

underlying expectancy (that affects compliance) may have been acquired 

over repeated exposures to the task situations or may have been 

deliberately evaluated during the initial exposures to the task. For 

unfamiliar tasks or for a goal level in a familiar task that is already 

high, raising the goal level further may detract from the motivation to 

comply. In such situations, the compliance route is likely to reverse 

its nature and acquire the property of a "rejection mechanism." That 

is, when expectancy is low, raising externally set goal levels will not 

contribute to increased intention to perform, and may even actually 

lower such intentions. 

Proposition 8: Expectancy of task success moderates the influence of 

internal representation of goal on intention to 
perform as follows. At high expectancies, higher 
levels of internal representation of goal contribute 
positively to intention to perform; while at lower 
levels that contribution is attenuated. At very low 
expectancy levels, the internal representation of goal 
may either have no influence on intended performance 
or may affect it negatively. 


Set Goal 



of Goal 


Intention to 




Task Success 

Figure 3.6: Schematic Representation of the Moderating Influence of 


Intention to 

High Expectancy 

oderate Expectancy 
Low Expectancy 

Internal representation 
of goal difficulty 

Figure 3.7: An Example of Interaction Between External Goal and 

Expectancy of Task Sucess 

External Goal Level and Perceived Extrinsic Rewards 

A very important characteristic of the salesperson's performance 
domain is that compensation is often tied to the performance level. 
Sales and Marketing Management (1986) reports that over 90% of firms 
selling consumer products and over 85% of firms selling industrial 
products employ compensation plans that are contingent upon performance 
in some manner. In such plans, performance levels are rewarded in the 


form of higher total compensation packages. Under such circumstances, 
higher levels of externally set performance goals would undoubtedly 
contribute to higher "perceived extrinsic rewards." This may be 
represented schematically in Figure 3.3 by drawing a link between 
"externally set goal" and "perceived extrinsic rewards." 

When the external rewards are designed to increase with the target 
performance level in this manner, it contributes to an accelerated 
effect on the goal setting phenomenon. Increasing externally set goal 
levels in this case not only raises performance through the direct 
compliance route, higher target performance levels also result in the 
raising of perceived extrinsic rewards, which further bolster the 
aforementioned effect. The widespread use of commissions as a 
motivational tool used by sales managers is consistent with this 
observation. A fuller exploration of this pathway, however reveals some 
important considerations and mediating factors that have significant 
implications for the proper use of this motivational tool. 

Types of Contingent Compensation Plans 

The actual compensation plan that a firm may decide to employ 
could be chosen from an infinite number of potential plans, each of 
which is contingent upon the performance dimensions in a different way. 
At one end of the range, an organization may choose a compensation plan 
that involves "straight salary"--a payment scheme in which the 
compensation is unrelated to sales performance (except in the context of 
long run evaluation and promotion) . At the other extreme, "straight 
commissions" are those schemes in which compensation is entirely 


dependent upon sales performance. In mixed compensation plans, salary 
refers to the fixed part of the compensation (which is unrelated to 
short-term sales performance) , while commissions refer to the amount 
earned directly as a result of the sales performance for the term. 

In addition to salary and commission, a compensation plan may also 
consist of a "bonus." While both commission and bonus are dependent 
upon sales performance, there are some important differences between 
them. Commissions are paid as an amount that correlates directly with 
the performance level achieved on some set of measures (e.g, sales 
volume, dollar sales, or profit margin) on a predetermined basis and are 
usually administered monthly. A bonus, on the other hand, is a 
discretionary payment for achieving a level of performance and is 
usually paid annually. While mixed plans consisting of both salary and 
commission are more common than both straight salaries and straight 
commissions, the most common combination plan includes salary, 
commission, and bonus (Dalrymple 1988, p. 385) . 

Sales organizations may also offer the salesperson the chance to 
purchase stock in the company at some future date at a preset price-- 
usually lower than the prevailing market value — in order to retain 
exceptional salespeople (Anderson, Hair & Bush 1988) . Such stock 
options, often referred to as "golden handcuffs," and similar payment 
schemes (e.g, "drawing accounts") exemplify the complexity that 
contingent payment schemes may involve. 

The criteria upon which sales performance is evaluated for the 
purpose of determining performance-contingent compensation may also take 
one of several forms, some of which are more complex than others. Such 


criteria may involve one or more performance dimensions. Among those 

dimensions that are frequently used as criteria for computing contingent 

payment, the most common indices include dollar sales, unit sales (or 

quantity sold) , profit margin achieved, number of sales call made, 

number of new accounts created, number of existing accounts serviced, 

and so forth. Most often, a combination of more than one of these 

indices are used with different weights attached to different criteria. 

In addition, when salespeople are responsible for more than one product 

line, different products may contribute differentially toward the 

individual's overall performance index, with a different set of criteria 

and weights used for each of the products. Again, to complicate matters 

further, additional differences may be involved across territories, 

seasons and markets. 

The preceding observations serve to justify the proposition that 

the relationship between target performance levels and perceived 

extrinsic rewards may be impeded by the difficulty involved in forming 

an impression of how clearly or strongly external rewards correlate with 

difficulty of goals. This is an important factor in the context of 

administration of performance-contingent compensation plans. It 

suggests that it is more effective to incorporate a reward structure 

that is perceived to be clearly related to (i.e., contingent upon) 

target performance rather than introduce complex compensation schemes in 

which the reward-performance contingencies are actually high, but are 

not perceived as such. Propositions 9 and 10 summarize these concepts. 

Proposition 9: In sales organization settings that incorporate 

performance-contingent compensation, higher levels of 
externally set performance targets contribute to 

Proposition 10 


increased perceived extrinsic rewards and contribute 
toward enhancing the goal setting effect. 

The characteristics of the compensation plan moderate 
the relationship between external goal and perceived 
extrinsic rewards as follows: 

(a) compensation plans involving a higher proportion 
of commissions (in the mix of commissions, salaries, 
and bonuses) result in stronger relationship between 
external goals and perceived extrinsic rewards; 

(b) complexity of contingent payment schemes attenuate 
the relation between external goals and perceived 
extrinsic rewards. 

Figure 3.8 schematically depicts the pathway between "external goal 
level" and "perceived extrinsic rewards" and the moderating effect of 
the contingency plan upon it . 

Contingency of 
Payment Scheme 






Figure 3.8: 

The Extrinsic Reward Pathway and the Moderating Effect of 
Contingent Payment Plan 

External Goal and Intangible Extrinsic Rewards 

Although the preceding sub-section inquiring into the issues 
concerning tangible rewards (embodied by financial compensation) 
addresses a crucial aspect of sales performance and motivation, the 
effect of intangible rewards on motivation in the sales force cannot be 
overlooked. In a survey of 121 senior sales executives, overall 
financial compensation was rated as only the sixth most effective of 17 
types of rewards for motivating salespeople (Miller 1979) . This study 


found "special recognition" for outstanding performance to be the top- 
ranked motivator and "encouragement and contact of supervisor" to be 
ranked third. 

In sales organizations, such forms of rewards may take various 
forms ranging from the subtle to the ostentatious and are often referred 
to as "recognition awards." Examples of these include trophies, wall 
plaques, bestowal of titles, e.g, "Salesperson of the Month" — a favorite 
of car dealers and real estate firms, publicity in house organs, 
congratulations (verbal and written) by senior executives, gifts for 
spouses of outstanding salespeople, memberships in honoraria such as 
"Super Salesperson Club," and so forth. Haring & Morris (1968) report 
that 90% of firms use honor awards and consider their effectiveness in 
motivating salespeople to range from "good" to "excellent." 

The common features that these rewards share with the previously 
enumerated financial rewards is that they are administered and 
controlled by the salesperson's supervisors and are contingent upon 
evaluations of the salesperson's performance. 

It is proposed that since high (or effective) levels of sales 
performance are almost universally desired by all sales supervisors, 
some form of perceived encouragement — however subtle — is associated with 
the reaching of high goal levels. Simple acknowledgements by 
supervisors of the salesperson's success in achieving set quotas may be 
sufficient for creating that perception. Also, individuals may assume, 
sometimes even below a conscious level, that high performance and 
achieving difficult performance goals inexorably portend recognition or 
praise by co-workers and supervisors. 

Proposition 11: High levels of performance targets are direct 

antecedents of high levels of perceived intangible 
extrinsic rewards. 

Moderating Influences on Intangible Rewards 

It has been proposed earlier that administrative factors 
(associated with the contingent payment scheme) may moderate the effect 
of external goal on perceived financial rewards. Similarly, 
administration factors may moderate the relationship between goal level 
and perceived intangible extrinsic rewards. 

Some firms actively pursue a policy of creating an organizational 

climate that is conducive to providing salespeople with encouragement 

and support. Several companies firmly believe that their 

"organizational culture" will contribute toward creating a work 

environment characterized by a high general level of awareness of the 

accomplishments of salespeople and other employees. It is believed that 

such a culture motivates employees to perform more effectively (Kelley 

1986; Templeton 1986) . In other words, organizational culture is an 

administrational factor that may have a moderating influence in the 

context of intangible extrinsic rewards. 

Proposition 12: Organizational culture moderates the relationship 

between external goal and perceived intangible 
extrinsic rewards as follows. For salespeople in 
firms with supportive organizational climates, the 
relationship between performance targets and perceived 
intangible extrinsic rewards will be stronger than for 
salespeople in organizations that do not have such 
climates . 

Figure 3.9 is a schematic representation of the moderating effect of 

organizational culture on perceived extrinsic rewards. 



X t 



Figure 3.9: The Moderating Effect of Organizational Characteristics on 

Perceived Extrinsic Rewards 

The Goal-Expectancy Pathway 

The preceding sub-section and the schematic representation in 
Figure 3.6 suggest that the external goal level and expectancy of 
success at the task are independent constructs that may contribute 
separately toward ultimate performance. While this presentation is 
useful for the purpose of elucidating the moderating effect of 
expectancy, the complete role played by the expectancy variable in the 
proposed model is somewhat more complex. 

As discussed in Chapter 2, several studies have recognized that 
there exists an inverse relationship between the level of external goal 
and the expectancy of task success. This suggests that the individual 
task performer recognizes that difficult goals imply that his or her 
probability of success at the task will be lower. 

A careful examination of the preceding implication reveals that 
individuals must somehow factor in their personal capabilities before 
they can arrive at estimates of expectancy of task success. For the 
purpose of our analysis, we may consider an individual's abstract 
representation of the assigned goal to be the primary antecedent of 

his/her interpretation of what the goal level means in personal terms . 
That is, after the assigned goal has been encoded at the general (and 
impersonal) level, this encoding is further processed with reference to 
the task performer's self -system; especially when the individual is 
confronted with the task. 

The higher the level of the internally represented goal, the 
higher will be the task performer's estimate of difficulty of the goal 
with r egard to personal standards . The personalized estimate of goal 
difficulty may be represented in terms of how much personal resources 
(effort and planning) would be demanded by the goal level associated 
with the task. That is, the "cognitive calibration" of the personalized 
estimate of goal difficulty occurs in terms of the level of effort and 
planning that the goal level imputes. Higher requirements of personal 
resources would then indicate lower expectancies of task success. 

The preceding idea embodies the "limited processing capability" 

view of the individual's cognitive system. If the goal difficulty is 

increased, then it cannot be functional for the individual to respond in 

any way other than to associate with it a lowered probability of task 

success . 

Proposition 13: Increased levels of internal representation of goal 

contribute to higher levels of the representation of 
the goal in personal terms, i.e., in terms of the 
level of personal resources that will be required to 
accomplish the goal. 

Proposition 14: The relationship between the level of personal 

resources required and the expectancy of task success 
is negative: higher estimates of personal resources 
required contribute to lowered expectancies. 


As a subjective estimate of the individual's probability of task. 

success for a given level of goal, expectancy is responsive to 

information pertaining to objective characteristics of the task and the 

goal level. The proposition that higher levels of externally set goals 

contribute to lowered expectancies is consistent with this viewpoint. 

It is therefore reasonable to expect that the assigner of the task may 

be able to provide the task performer with other sources of relevant 

information (in addition to the actual goal level and task) that will 

also influence expectancies of task success. For instance, independent 

assessments of the individual's chances of task success (by the assigner 

of the task or another external agent) may also influence the composite 

expectancy that is formed by the task performer. The source credibility 

associated with the supplier of the information, however, should 

moderate how strongly these external estimates affect the individual's 

expectancy of task success. 

Proposition 15: Independent external estimates of the individual's 

probability of success influence the person's 
(internal) expectancy of task success. 

Proposition 16: The credibility associated with the external source 

will moderate how strongly the external estimate of 
probability of success influences expectancy: high 
source credibility will contribute to stronger 
influences of independent external estimates 
(information) on expectancy than low credibility 
sources . 

The preceding propositions (13 through 16) are consistent with the 

view that expectancy is influenced by objective characteristics of the 

task and other sources of external information. Additionally, they 

identify the cognitive processes that intervene in the influence of 

objective task factors on expectancy. 


Figure 3.10 overlays the Goal-Expectancy pathway on the central 
Extended Compliance pathway. Propositions 13 and 14 are represented by 
the links between "internal representation of goal difficulty, " 
"estimate of personal resources required," and "expectancy." 
Proposition 15 is embodied by the link between "independent external 
estimates" and "expectancy"; and Proposition 16 is schematically 
depicted by the moderating effect of source credibility. 

Set Goal 

Representation of Goal 


Intention to 

/ * 


Expectancy of 
Task Success 

Estimate of Personal 
Resources Required 




Source Credibility 



External Estimates 

Figure 3.10: The Goal-Expectancy Pathway Overlaid on the Extended 

Compliance Pathway. 

Figure 3.10 also clearly elucidates how increasing levels of 
externally set goals may affect performance in two opposing ways at the 
same time. Raising external performance targets normally enhance actual 
performance through the compliance path. Concurrently, increased goal 
levels lower expectancies which, in turn, can slow down or even reverse 
this positive influence. The following sections of this chapter will 
further examine these contingencies and other pathways involved in the 
processes . 

Nature of the Goal-Expectancy Relationship 

Although increasing externally assigned goals have the effect of 
diminishing expectancies of task success, there is no reason to believe 
that expectancy decreases proportionally with increasing goal levels. 
That is, although the relationship is monotonically negative, it may not 
be linear. In fact, it is likely that the curve depicting the 
relationship between external goal level and expectancy will consist of 
some characteristic elbows (or turning points) that will have 
significant effects on the manner in which expectancy moderates the goal 
setting effect. The nature of the shape of this curve is described as 
follows . 

To begin to chart the representation under consideration, we begin 
at the very low end of goal difficulty and examine how expectancy 
changes in response to increases in goal difficulty. When external goal 
levels are very low (close to zero perhaps) , increases in the level of 
external goal have little (or negligible) effects on expectancy of 
success. The downward sloping curve (representing diminishing 
expectancy in response to increasing external goals) is relatively flat. 
This is depicted by the segment AB in Figure 3.11 (which describes the 
shape of the goal-expectancy curve over an extended range) . 

At these low levels of goals, the task performer is likely to 
perceive no significant changes in difficulty levels with regard to 
his/her capabilities as the physical goal level is increased. As an 
example, a competent athlete would consider the task of jumping over a 
two feet high bar to be only marginally more difficult, if at all, than 
jumping over one that is only one foot high. Similarly, a competent 

of Task 

Level of External Goal 
(Internal representation of goal difficulty) 


Figure 3.11: The Relationship Between Goal Level and Expectancy of Task 


salesperson (who realizes several thousand dollars worth of profit 
through sales each year) would perceive very little difference in task 
difficulty in the course of confronting an annual sales quota of $100 
and another of $200. 

It is within this "zone of confidence," represented by AB in 
Figure 3.11, that the effect of increasing goals on compliance will be 
most marked. Because of the very way in which goals are perceived, the 
compliance mechanism will work unhindered by lowered expectancies as 
goals are increased. That is, as goal levels are raised, the direct 
effect on "intention to perform" will be relatively undisturbed by the 
marginal decreases in the level of expectancy of task success. 

As the physical level of the external goals are raised, the 
individual will sooner or later cross some threshold level of goal and 
reach a "zone of uncertainty." At these goal levels, the personal 
resources required to confront the task are perceived to approach the 
individual's own maximum resource levels. Raising goals further would 


correspond to much sharper drops in expectancy estimates. In this zone, 
represented by the segment BC in Figure 3.11, the expectancy estimates 
will be very sensitive to changes in external goal levels. The 
threshold goal level corresponding to the beginning of the zone of 
uncertainty is represented by Tl. While this may occur at different 
levels of external goals for different individuals, it is proposed that 
the shape of the curve depicts a basic characteristic in the way 
expectancy estimates will respond to changes in the external goal 
levels . 

Finally, as external goals are raised even further, the task 
performer will cross the upper threshold Ty and enter another zone of 
relative certainty, BC. Goal levels in this zone correspond to those 
that are perceived to be equivalent to the impossible (e.g., jumping 
over a 30 feet bar, or realizing a profit of hundreds of millions of 
dollars, etc., in the context of the earlier illustrations) . 

In the zone of uncertainty, increases in external goals rapidly 
lower expectancy estimates. Consequently, the effect on performance 
(through the compliance mechanism) suffers on account of the moderating 
effect of lowered expectancy. When the expectancy falls below some 
critical level, the compliance mechanism can be expected to reverse to a 
"rejection mechanism" and further increases in external goals would be 
likely to lower intention to perform. 

The preceding observations suggest that for a certain range of 
goal difficulty, particularly those characterized by the zone of 
confidence, one would observe a monotonic positive relationship between 
goal difficulty and performance. This is in accordance with the 

observations of goal theorists. In this region, as some researchers 
have proposed, goals will be "accepted" by the task performer. Others 
merely suggest that "goal acceptance" or "goal commitment" is likely to 
be high. In any case, empirical results evincing positive monotonic 
relationships could be accounted for. 

Empirical evidence negative monotonic relationships between goal 
levels and performance are also accommodated by the preceding conceptual 
organization. Specifically, when the "compliance mechanism" is reversed 
at extremely low levels of expectancy, such negative effects will be 
observable . 

Finally, the transition from the positive monotonic to the 
negative monotonic slope (as goal difficulty levels are raised from the 
very low to the very high) generate an inverted-U shaped curve, evidence 
for which exists in literature as well. 

The theoretical perspective offered here suggests a dynamic 
interplay between some external variables and intermediate cognitive 
processes that naturally generate different kinds of goal-performance 
relationships. In the process, it offers insight into how the 
apparently conflicting empirical observations can be systematically 
explained and interpreted in the context of an organized framework. 
Also, the preceding framework is consistent with Goal Theory and 
Expectancy Theory postulates. 

Earlier attempts at explaining inverted-U curves have referred to 
Atkinson's Achievement Motivation Model, since it predicts inverted-U- 
shaped relations between "subjective probability of success" and 
"tendency to succeed." However, it has typically been positioned as a 

8 9 

rival paradigm positioned as a rival paradigm to both Expectancy Theory 
and Goal Theory. It requires to be mentioned that, in addition to the 
inverted-U-shaped curve, Achievement Motivation Theory also predicts 
flat and U-shaped curves under certain conditions. 

The rationale presented in this section purports to be a more 
complete explanation of the bell-shaped goal-performance relationship 
than those that invoke the Achievement Motivation theory simply because 
it includes a similar prediction in a different setting that involves 
certain specific and restrictive assumptions regarding individual 
difference variables. 

The theoretical perspective of this discourse is not aimed at 
establishing that the Achievement Motivation Model is invalid or is 
inconsistent with the existing evidence. In simply suggests that the 
observed inverted-U-shaped relationships do not automatically "disprove" 
Goal Theory and Expectancy Theory postulates (and establish a competing 
viewpoint) , but are natural extensions of both. At the same time, they 
may be in accordance with any number of models that include inverted-U- 
shaped relationships (between goal and performance) among their 
predictions . 

Proposition 17: The negative relationship between external goal and 

expectancy of task success is nonlinear. 

(a) at low levels of goal difficulty, overall 
expectancy is high but the decrease in expectancy in 
response to increase in goal levels is small; 

(b) at very high levels of external goals, overall 
expectancy is very low and again changes in goal level 
produce small responses in expectancies; and 

(c) at intermediate levels of external goals, the 
expectancy level decreases sharply in response to 
increases in external goals. 


Proposition 18a: At low levels of external goals, the compliance 

mechanism is predominant: increases in external goal 
levels contribute to increased intention to perform, 
actual resources expended, and performance. 

Proposition 18b: At very high levels of external goals, increases in 

external goal levels affect intention to perform, 
actual resources expended, and performance negatively. 

Proposition 18c: As externally set goal levels are increased from the 

very low to the very high, the relation between the 
goal level and intention to perform (as well as 
between goal and performance) exhibits an inverted-U- 
shaped relationship. 

The conceptual framework described in this chapter identifies some 
more pathways involved in the effect of goals on performance. These 
pathways shed additional light on why different kinds of goal- 
performance curves may be generated under differing circumstances and 
are examined in the following sections. 

The Moderating Effect of Self-Ef f icacv 

The preceding section enumerates processes involved in the 
influence of external sources of information on the expectancy variable. 
However, there may also exist influences upon expectancy which originate 
from within the individual's cognitive system. Earlier studies have 
proposed that self-esteem and other measures of the individual's own 
appraisal of the self influence expectancy (see Chapter 2 for a fuller 
description) . 

In this section, it is proposed that self-efficacy has an 
important and significant moderating effect on the formation of 
expectancy. Since self -efficacy refers to the individual's sense of 
mastery in the given task domain, those who are low in self -efficacy are 
likely to perceive their personal chances of success at a given task 


task differently than those high in self-efficacy. Self-efficacy, 
however, is not proposed to influence the formation of the abstract 
cognitive (or internal) goal representation, since that encoding is 
believed to occur before self-related cognitions are generated. 

On the other hand, the effect of the abstract "internal 
representation of goal" on the "estimate of personal resources required" 
invokes the individual's evaluation of personal standards and appraisal 
of personal resources available in light of the task goal as perceived 
by the individual. It is this relationship that is moderated by self- 
efficacy. This moderated effect is also reflected in the resultant 
level of expectancy. Eventually, self-efficacy affects "intention to 
expend effort" as well as performance through its enhancing effect on 
expectancy. Figure 3.12 depicts the preceding interactions 

Set Goal 

Representation of Goal 

Intention to 


Estimate of Personal 
Resources Required 


Expectancy of 
Task Success 

Figure 3.12: Moderating Effect of Self-Efficacy on Estimate of Personal 

Resources Required 

Proposition 19: 

Self-efficacy moderates the effect of "internal 
representation of goal" on "estimated personal 
resources required" (and consequently on expectancy) 
as follows: 


(a) for individuals high in self-efficacy, increased 
levels of internal representation of goal contribute 
to smaller, increases in estimated personal resources 
necessary than individuals low in self-efficacy; and 

(b) for individuals high in self-efficacy, increased 
levels of internal representation of goal contribute 
to smaller decreases in expectancy than individuals 
low in self-efficacy. 

The Intrinsic Motivation Pathway 

In the process of attempting to account for the goal setting 
effect, several studies have espoused the notion that difficult task 
goals are perceived as more "challenging" by the task performer than 
easy goals. This perception of challenge is believed to energize the 
task performer toward higher levels of performance. In this section of 
the chapter, issues that are of significance to the underlying cognitive 
processes implicated by this idea are examined and the contingencies 
associated with them are explored. 

The Concept of Intrinsic Reward.? 

One of the distinctive features of the "challenge" (that the 
individual task performer is believed to perceive in difficult goals) is 
that it is a reward— a desirable consequence of undertaking the task 
performance— which is not administered or controlled by an external 
agent. There may exist other such desirable (or positively valenced) 
feeling states associated with difficult goal levels, such as 
satisfaction, enjoyment, fun, etc., all of which are internally 
generated and contingent upon the the individual's participation in the 
task. These rewards— generated internally by an interaction of the task 


performer and the task itself— may be classified as intrinsic rewards 
and are consistent with the definition stated earlier. 

Since intrinsic rewards are not administered by an external agent, 
it is possible to view them as features or characteristics of the task 
itself. For instance, it is not uncommon to describe an activity as a 
"challenging task" or as "satisfying work." The fact that the challenge 
or the satisfaction at the task depends in part upon the person involved 
in the activity indicates that it is more useful and less confusing to 
treat intrinsic rewards as separate from task features. Nevertheless, 
it requires to be mentioned at this point that this concept involves 
some amount of difficulty and, to the task performer, intrinsic rewards 
may often appear to be phenomenologically indistinguishable from the 
activity itself. 

Another important point in this context is that tasks or task 
goals may also contribute to negative intrinsic rewards depending on the 
individual characteristics of the task performer. This concept will be 
explored further and integrated with the overall intrinsic motivation 
pathway in a following section. 

The Rationale for Exte rn ally Administered Rewards 

The value of externally mediated rewards have been recognized in 
psychological and management literature for quite some time. Taylor 
(1911) recognized the effectiveness of piece-rate payments and wage 
incentives in his treatise on the "scientific management principle." 
Behaviorists such as Thorndike (1932) viewed behavior as something that 
could be evoked or shaped with the help of external reinforcements. The 


theoretical assumption underlying such outlooks of motivation is the 
individual will have an avoidance orientation with regard to task- 
relevant behavior. It is believed that the function of external reward 
is to create a source of attraction for the task performer whose 
positive valence (i.e., desirability) will outweigh the disinclination 
associated with the task performance. The residual valence (the 
positive valence of reward minus the negative valence of undertaking 
task activity) would constitute the resultant motivating force for the 
action. This implied or underlying assumption regarding work motivation 
and extrinsic rewards is evident even in contemporary term for financial 
payments, viz., compensation. Salaries, commissions, bonuses, and other 
payments are considered to be means for "compensating" a loss or a 
sacrifice borne by salespersons and other organizational employees. 

Relatively modern theories of work motivation (e.g., the 
Expectancy-Valence theory) also incorporate in part this assumption 
regarding work motivation.! Expectancy-Valence theory is partly based 
upon the implicit assumption that rewards (payments, reinforcements, 
etc.) are positively valenced while expenditure of effort at a task is 
negatively valenced. 

Mitchell (1974) observes that Vroom's (1964) Expectancy Theory 
postulates are designed to address the valences and motivational effects 
of externally administered rewards, not intrinsic rewards, although many 

Actually, a large number of similar models of the expected 
utility form (Tolman, Lewin, Edwards, Atkinson, Rotter, Vroom, Peak, 
Rosenberg, Dulany, and Fishbein) share this basic premise at some level 
or another. 


researchers have overlooked this. He also indicates that Expectancy 
Theory may be seen as based upon some sort of principle of maximization 
of expected pleasure under the constraints of positive valences of 
rewards and the disutility of effort. 

A closer examination of the concept of disutility of effort 
suggests that this disutility (or disinclination) associated with a 
given level of personal effort is actually a sub-set of the category 
comprising of negative intrinsic rewards. In other words, the class of 
motivation theories based upon the principle of subjective expected 
utility maintain the viewpoint that individuals associate high levels of 
anticipated personal effort as intrinsically less satisfying. That is, 
with all other factors remaining the same, higher effort levels should 
contribute to lowered intrinsic rewards (which by definition is 
inclusive of the effects of disutility) . 

Limitations of the Expected Utility Viewpoint 

In his model, Vroom does not address the issue of intrinsic 
rewards. However, he was aware of the possibility of their effects as 
is evident in his conjecture that "people may seek to do well in their 
job even though no externally mediated rewards are believed to be at 
stake" (Vroom 1964, p. 16). In fact, Locke's notion (that higher goal 
levels contribute to a heightened sense of "challenge") cannot be 
properly examined unless one explores the possibility that higher 
anticipated effort levels may contribute to increased intrinsic interest 

The intrinsic reward orientation may operate within or without the 
context of externally set goals. Pittman, Emery & Boggiano suggest that 


when an individual adopts an intrinsic motivational orientation, 
features such as novelty, complexity, challenge, and the 
opportunity for mastery experiences are sought and preferred. 
These qualities are usually present in some form during enjoyable 
play, entertainment, or leisure time periods. [1982, p. 790] 

Research dealing with these issues have recognized that there may often 

exist situations in which individual task performers will value (i.e., 

evaluate positively) increased effort at a task, simply because the 

activity itself is a source of pleasure. 

Relation Between External Goal and Intrinsic Rewards 

In cases when increased anticipated effort contributes to a higher 
level of dissatisfaction (the expected utility formulation) , as well as 
when higher levels of effort are valued positively, it is the 
anticipated levels of personal resources (effort) that directly 
influence intrinsic rewards. 

In an earlier section, it was proposed that increasing externally 
set goals contribute to to higher levels of internal representation of 
the goal, which in turn raises estimates of anticipated personal 
resources associated with the task. This estimate of personal resources 
necessary influence intrinsic rewards either positively or negatively 
depending upon a proposed moderating factor, as described in the 
following paragraphs. 

Intrinsic rewards have been found to be associated with tasks in 
which the individual perceives to possess (or is made to believe that he 
or she possesses) competence and mastery (Deci 1971; Pittman, Emery & 
Boggiano 1982; Pittman, Davey, Alafat, Wetherill & Kramer 1980) . Thus, 
in a specific task domain in which an individual possesses heightened 
self-efficacy, higher external goals are likely to contribute to 


increased intrinsic rewards. This happens because the individual 
evaluates complexity and effort positively in task domains in which he 
or she perceives to be personally competent. In case of individuals 
with low self-efficacy, increasing externally set goals will contribute 
to increased estimates of personal effort which is negatively evaluated. 
Therefore, higher levels of external goals will detract from intrinsic 
rewards (i.e., contribute negatively to intrinsic rewards) for those low 
in self-efficacy. 

Finally, intrinsic rewards — evident in individuals with high self- 
ef f icacy--may be further enhanced by controlling some key aspects of the 
task environment. Usually, individuals respond positively to task 
situations where the potential for immediate positive feedback is high. 
Actual feedback, if positive in nature and delivered promptly, aid in 
enhancing self -efficacy in the long run. However, increased intrinsic 
interest is achieved merely if the task performer perceives that there 
exist opportunities for obtaining prompt feedback regarding the worth or 
effectiveness of the effort expended. 

As an illustration, consider the increased interest one may have 
in approaching a test (or a game of skill or any such activity) if one 
is informed that the results would be immediately reported, in 
comparison with a similar activity where the results are to be made 
available after a month. 

Proposition 20: Estimates of personal resources required at a task 

influence the intrinsic rewards associated with it and 
is moderated by self-efficacy as follows: 
(a) for individuals high in self-efficacy, high levels 
of estimated personal resources required contribute to 
increased intrinsic rewards; 

Proposition 21: 


(b) for those low in self-efficacy, increased levels 
of estimated personal resources detract from intrinsic 
rewards . 

Tasks that are so designed that they provide quicker 
feedback with regard to performance are conducive 
toward generating a stronger and more positive 
relationship between "estimated personal resources 
required" and "intrinsic rewards." 

Task Characteristics 
(Feedback opportunity) 

Estimate of Personal 
Resources Required 





Figure 3.13: Moderating Effects on the Relationship Between Anticipated 
Personal Resources and Intrinsic Rewards 

Effect of Intrinsic Rewards on Performance 

Similar to extrinsic rewards, higher levels of intrinsic rewards 
contribute to increased "intention to perform." However, in one 
respect, there exists a significant difference in the way that effect 
occurs. Previous propositions suggest that extrinsic rewards and 
expectancy of task success combine interactively to enhance the 
compliance mechanism. Their effects may be thought of as combining 
multiplicatively, so that if the level of expectancy is zero, their 
combined effect is zero as well. In such a case, no amount of extrinsic 
rewards should motivate a person to undertake a goal level that he/she 
is absolutely certain fail in. Similarly, if we consider a task 


situation in which the extrinsic rewards are zero, even certainty of 

success would fail to motivate a person to undertake that activity, 

since no incentives exist. 1 

In contrast, intrinsic reward contribute additively toward 

intention to perform. They generate internal incentives for approaching 

and persisting at a task, sometimes even when an externally set goal has 

been surpassed. Everyday examples include intrinsically motivated 

individuals whose satisfaction with their work propel them to continue 

working for longer than that which is considered necessary or 

sufficient . 2 

Proposition 21: When an externally set goal level is within reach, 

individuals motivated by intrinsic rewards may 
continue to perform at the task even when the 
externally goal level has been met. In absence of 
intrinsic rewards, individuals will either just meet 
the target performance or abandon the task earlier. 

Figure 3.14 overlays the intrinsic motivation pathway on others. The 

output from intrinsic rewards is shown to contribute directly to 

"intention to perform," suggesting the additive effect of intrinsic 

rewards. In order to maintain clarity, the figure does not include the 

"extrinsic reward pathway." 

T In this case, it must also be assumed that no intrinsic rewards 
operate either. 

2 This example merely helps to illustrate the point. It is 
possible that an individual may undertake similar activities in 
anticipation of intangible external rewards such as recognition or 
similar bestowals. 


Set Goal 

Representation of Goal 

Intention to 


t ' 


Expectancy of 
Task Success 



— w ' X 



Estimate of Personal 
Resources Required 




Task Characteristics 

w s 


W '^ 


Figure 3.14: The Intrinsic Reward Pathway Overlaid on Other Pathways 

The Composite Effect of Self -Efficacy 

Figure 3.14 identifies self-efficacy as an important moderating 
variable on account of its influence on intention to perform through at 
least two different pathways. Higher levels of self-efficacy serve to 
enhance expectancy of task success (which bolsters the compliance 
mechanism) as well as increasing intrinsic rewards that add to the level 
of intention to perform. 

It may be proposed that individuals below a certain threshold 
level of efficacy are likely to respond negatively to increased levels 
of externally set goals and actually lower intention to perform when 
confronted with increasing target performance levels. This is because 
the increase in externally set goal levels would not only lower 
expectancy level below the critical but contribute to negative intrinsic 
rewards (or increased "disutility") . This is similar in nature to the 
idea of the reversal of the "compliance mechanism" described earlier. 
On the other hand, individuals with high self-efficacies may be expected 


to respond positively to increases in goal levels in accordance with the 

familiar goal setting effect. 

Proposition 22: The self-efficacy of the individual task performer 

interacts with the level of externally set goal as 
follows : 

(a) individuals who are high in self-efficacy respond 
to increased levels of externally set goals by raising 
their intention to perform (and consequently achieve 
high performance levels) ; 

(b) individuals low in self-efficacy may have lowered 
levels of intention to perform when confronted with 
increased levels of externally set goals; and 
consequently their performances are likely to suffer 
as well. 

This interactive effect of self-efficacy and externally set goal 

is considered significant for a number of reasons. First, it may help 

to account for the anomalies in existing empirical evidence and explain 

both the positively-sloped and the negatively-sloped goal-performance 

curves that have been reported in literature. Secondly, both goal level 

and self-efficacy are easily identifiable and measurable variables. 

While the external goal level is a direct exogenous variable and can be 

manipulated at various levels without much difficulty, individual's 

self-efficacies may also be influenced through "training programs" and 

educative measures in which the perceived mastery in the task domain of 

interest is gradually built up. Thus, from a theoretical as well as a 

practical standpoint, this interaction is of crucial importance in the 

conceptual model . 

The Static Model — Summary and Implications 

Each of the preceding sections of this chapter addresses a part of 
the overall conceptual framework that identifies the different cognitive 
mechanisms involved in the effect of the level of externally set goal on 


performance. This conceptual model is referred to as the Static Model 
since it examines how and why the goal setting effect is generated 
without taking into explicit consideration the possible dynamic 
interactions among some of the moderating variables when external goals 
are assigned repeatedly over a period of time. The following section of 
this chapter inquires into some of the issues relevant to the dynamic 
effects of goal setting. 

In sum, the static conceptual framework suggests that the level of 
externally set goals influence performance through several interactive 
mechanisms, since the increase in external goal level affects a number 
of related cognitive dimensions. First, an individual usually responds 
adaptively to an increased level of an external goal merely in order to 
comply with the task requirements. This compliance in turn is 
influenced by the perception of increased extrinsic rewards, which are 
signalled by higher goal levels themselves. Higher goal levels also 
contribute to lowered expectancies which again serve to dampen 
compliance. However, for those individuals who are high in self- 
efficacy, this negative influence of the increase in goal level takes 
place at a slower rate than those low in self-ef f icacy . Finally, 
increased goal levels may contribute to or detract from intrinsic 
rewards depending upon the characteristics of the task and the 
individuals' self -efficacy level. High levels of intrinsic rewards 
contribute additively (i.e., independently and separately) to the 
individual's intention to perform at a higher level. 

This model identifies a number of interactive cognitive variables that 
intervene in the goal setting effect and also enumerate the relevant 


exogenous variables that are of significance in the context of 
administration and control. Figure 3.15 depicts the several interactive 
pathways. The exogenous and controllable variables which are identified 
as important moderators in the process are highlighted in the figure 
(and are shown as enclosed in rectangles with bold borders) . An 
important cognitive variable identified in this framework is self- 
efficacy. It is proposed that self-efficacy is a concept of consider- 
able significance and is likely to contribute in our understanding of 
the issues surrounding goal setting. 

Dynamic A spects of Goal Setting 

The conceptual framework dealing with static issues (i.e., those 
which do not take into account the effects of repetition and feedback) 
identify a number of variables that are of importance in the goal 
setting process. Included among those is self-ef f icacy which is 
proposed to interact with the level of externally set goal to generate 
some significant effects. In other words, externally set goals and 
self-efficacy were treated as orthogonal independent variables in the 
conceptual model. In theoretical treatments of the repeated effects of 
goal setting, however, it is possible to observe that the manner in 
which external goals are scheduled may influence self-efficacy and 
ultimately influence the performance level that can be achieved. 

Proxim al and Distal Rnal.s 

Bandura & Schunk (1981) defined "goal proximity" as that property 
of goal that indicates how close (i.e., near or "proximal") the deadline 



















00 9 



a a 


o z 

a) O 









r ) 






for achieving the target performance level is to the individual task 
performer. For instance, a target performance level that must be 
achieved in a month's time may be considered more proximal than an 
annual quota for a salesperson. However, this goal is also more 
"distal" (i.e., further removed in time) than a weekly quota. 

According to these authors, goal proximity "is especially critical 
the more closely referential standards are related to ongoing behavior, 
the greater the likelihood that self-influences will be actuated during 
the process" (Bandura & Schunk 1981, p. 586) . Bandura & Simon (1977) 
and Jeffery (1977) also provide some indication that the impact of goals 
on behavior may be determined by how far into the future the goals are 

Based on these considerations, these researchers indicate that the 

goal setting phenomenon can be made even more effective if distal goals 

are replaced by proximal sub-goals . For instance, an annual sales quota 

of $120,000 worth of profits may be replaced by monthly quotas of 

$10,000 on an average. 1 Such proximal sub-goals are believed to enhance 

performance on account of their influence on self-efficacies and 

intrinsic interest. The authors also suggest that 

proximal sub-goals can . . . serve as an important vehicle in the 
development of self-percepts of efficacy. ... By contrast, 
distal goals are too far removed in time to provide sufficiently 
clear markers of progress along the way to ensure a growing sense 
of self-efficacy" (Bandura & Schunk 1981, p. 587) . 

•'■Different months may be apportioned different quota levels in 
consideration of seasonalities, if any. 


In an experimental setting, these researchers found evidence for 
the superior effect of proximal sub-goals on performance. Additionally, 
the perceived self-efficacy generated by the use of proximal sub-goals 
was found to be positively related to intrinsic interest in the task. 

Although the aforementioned study experimentally tested the 
effectiveness of proximal sub-goals under conditions of self-directed 
learning (in which the goals were self-set in accordance with the 
experimenter's "suggestions"), the theoretical ramifications have 
implications for task situations in which performance goals are 
externally set. 

Dual E ffects of Proximal Goals 

The preceding study is consistent with the static conceptual model 
described earlier. In that model, it is suggested that self-efficacy 
influences performance through its enhancing effect on expectancy which 
in turn raise intention to perform. Self-efficacy is also believed to 
raise intrinsic rewards which contribute to intention to perform as 
well . 

The conceptual model, however, sheds light on another effect that 
externally set goals may have. It was mentioned earlier that external 
goals have attention enhan cing properties . The attention enhancement 
associated with externally set sub-goals that are repeatedly assigned 
aids in maintaining a stronger internal representation of the goal over 
the entire period for which the target performance is assigned. 
External goals that are assigned in task situations in which they may 
signal extrinsic rewards (as in the salespersons' work domain) will also 


bring with them these properties even when they are administered as sub- 
goals . 

The static models thus lends an additional important justification 
for the results observed in the study by Bandura & Schunk (1981) . This 
theoretical observation would not have been important if the repeated 
attention enhancing properties of sub-goals and its positive effect on 
self-efficacy contributed to "intention to perform" (and performance) in 
the same way under all circumstances. The following scenario 
illustrates how these two factors may have differential contributions on 

If each (or most) of the proximal sub-goals are associated with 
failure experiences (because of either the task characteristics, goal 
level or individual characteristics) , repeated failures will contribute 
to a reduction in (i.e., lowering of) self-efficacy levels instead of 
enhancement. In such a case, a distal goal would not have as strong an 
effect in terms of self-efficacy as proximal sub-goals and should appear 
as a relatively more effective tool compared to proximal sub-goals. 
However, the proximal sub-goals will continue to benefit from the 
repeated exposures to the attention enhancement properties even in 
failure situations. It is proposed that under such conditions, proximal 
sub-goals may remain more (or less) effective than distal goals 
depending on which of the two effects are greater. 

Proximal sub-goals are usually associated with multiple feedbacks 
regarding performance levels achieved (either administered deliberately 
by external agents or derived directly from observation of task results 
by the individual task performer) . Each time a sub-goal is set, the 


individual has access to information regarding success or failure 
(usually success), which contributes toward developing self-ef f icacy . 
That is, while proximal sub-goals provide repeated attention enhancement 
as well as feedback opportunities, a distal goal provides neither. 


As is evident, the knowledge of the nature of the effect of 
proximal sub-goals (obtainable by using the experimental design 
suggested in the preceding sub-section) have important ramifications for 
the design of quota schedules. For instance, quotas are perhaps best 
scheduled on a proximal basis, in general. However, if the 
salesperson's initial career or a specific territory involves some 
setbacks or failures, a distal quota (such as an annual goal with 
corresponding appraisals) may be helpful. Additionally, sub-quotas for 
periods of smaller duration may be provided for the purpose of 
"reminding" the individual, without administering the usually 
accompanying negative feedbacks. 

For tasks that are relatively easy (i.e., successes are most 
likely) the relatively lower administrative costs of setting long term 
quotas (over short term quotas) may be preferred; but motivation and 
performance may be enhanced by providing feedbacks about successful 
achievements at regular intervals. 

For tasks in the range of moderate difficulty, associated with 
equal chances of success or failure (i.e., tasks associated with 
"challenge") proximal sub-goals in their pure form would be recommended 


as it would bolster self-efficacy after initial successes as well as 
allowing for the benefits of repeated attention enhancement. 


This chapter describes the methodology used for testing a number 
hypotheses in order to corroborate some of the key propositions outlined 
in the preceding chapter. It also provides evidence in support of the 
validity of the scale developed specifically for measuring self-efficacy 
in the context of negotiation tasks. This scale has been used to 
measure self-efficacy so that the hypotheses (concerning externally set 
performance goals and measured self-efficacy) could be tested within the 
context of an experiment which involves a negotiation task setting. 

The conceptual model and the theoretical propositions described in 
the previous chapter address broad issues whose relevance is not limited 
merely to the sales force. Some parts of the model, however, have dealt 
with concerns that are significant only within a sales force context. 
The experimental setting described in this chapter, on the other hand, 
is adapted to closely match the environment of sales activities so that 
measures of performance, goals, and other variables that are especially 
characteristic of sales situations may be included. 

Figure 4.1 depicts the reduced model which incorporates 
essentially those elements which will be tested in order to provide 
empirical support for some of the key constructs and relationships 



outlined in the previous chapter. Thus the reduced model embodies only 
the critical constructs of the overall static model (Figure 3.15) and 
identifies the interrelationships that are of significant interest from 
a theoretical standpoint. 








Figure 4.1: The Reduced Model 

Essentially, the reduced model identifies three important indepen- 
dent variables: the level of externally set goal, the self-ef f icacy of 
the individual performing the task, and external information regarding 
the task that is provided to the task performer. The two intervening 
variables of primary importance are "expectancy" and "intrinsic 
interest." The dependent variable of crucial theoretical interest is 
"effort expended" by the task performer. 


The reduced model indicates that the effort expended in the 
context of a sales related task is primarily dependent upon the target 
performance level (i.e., the level of externally set goal). Represented 
schematically by the horizontal pathway between "external goal level" 
and "intention to expend effort, " this proposed relationship embodies 
the "compliance relationship" explained in detail in the preceding 
chapter. The expectancy of task success is inversely related to the 
level of externally set goal and moderates the compliance pathway. 
Additionally, the goal-expectancy relationship is moderated by self- 
efficacy. Self-efficacy also moderates the impact of the level of the 
externally set goal on effort expended via the intrinsic reward pathway. 
External information regarding the task, particularly that which 
pertains to the level of difficulty, is also proposed to be a 
determinant of of the expectancy of task success. 

In forthcoming sections of this chapter which deal with experi- 
mental design and development of hypotheses, the theoretical proposi- 
tions embodied by the reduced model are represented in terms of 
operational hypotheses within the context of the task chosen for the 

The following section introduces the general experimental paradigm 
selected and provides a rationale for the choice. Later sections 
describe the study in detail and enumerate the experimental hypotheses 
within the context of the task selected for the study. These experi- 
mental hypotheses are merely operational transcriptions of key theoreti- 
cal propositions delineated earlier, and represent them in the setting 
of an experiment. 

The Experimental Paradigm 

The Task Setting 

The performance domain of salespeople that is of interest in this 
study is one that is characterized by target performance levels or sales 
quotas (that are typically assigned in accordance with overall sales 
forecasts) , competitive performance domains (often involving 
negotiations with regard to the terms of sale) , and certain amounts of 
uncertainty regarding clients' interests, power and ability. 

In view of the aforementioned considerations, the performance task 
selected for the experimental setting involves a bargaining scenario. 1 
It is in the context of this bargaining task that the performance 
targets are assigned. The actual performance level achieved (and other 
dependent, independent, and intervening variables) are also measured 
within the context of the same task. 

Advantages of a Microcomputer Based Experimental Methodology 

As a part of the experimental procedure, the subjects are each 
assigned the role of an industrial salesperson representing a specific 
company that has a select set of clients to whom the salesperson may 
sell the product. The salesperson is assigned the task of negotiating 

1 Although negotiated terms of sale are more prevalent in the work 
domain of salespeople, research indicates that some important consumer 
transactions are also associated with bargaining as well (Johnston & 
Bonoma 1983) . Pennington (1968) identifies shopping in automobile 
showrooms and flea markets, durable goods shopping, and in-home 
transactions as consumer activities that are typically accompanied by 
bargaining behavior. 


the terms of the sale with the buyers. The roles of the buyers are 
handled by a simulation program on a microcomputer. Effective 
negotiation (involving persistent and motivated bargaining) will be 
reflected in the terms of sale finally settled upon by the salesperson 
(subject) and the buyers (represented by a computer program) . 

This laboratory setting permits strict control over key 
independent variables to be manipulated in the experiment and allows for 
efficient measurement of relevant dependent and intervening variables. 
It thus facilitates in the making of a stronger statement regarding 
causality . 

Since the performance task selected involves bargaining between 
salespeople and buyers, this study shares some common concerns and 
features with other studies on negotiation in organizational buying and 
channels behavior (e.g., Clopton 1984; McAlister, Bazerman & Fader 1986; 
Neslin & Greenhalgh 1983; Schurr & Ozanne 1985; Stern, Sternthal & Craig 
1973) . Previous studies on buyer-seller relationships have made some 
important recommendations regarding how the inquiry into these and other 
related issues should be conducted. For instance, conceptual models of 
the selling process identify the buyer-seller dyadic interaction as one 
that involves several complexities and espouse the use of experimental 
settings rather than using static survey methods (Spiro, Perrault & 
Reynolds 1977; Weitz 1979, 1981) . Additionally, Clopton & Barksdale 
(1987) make another key suggestion. They indicate that earlier 
experimental studies (e.g., Komorita & Barnes 1961) typically relied on 
"paper and pencil" methodologies involving role playing by both the 
parties in the bargaining activity. They observe that most of these 


studies that incorporate a two-party role playing format but actually 
inquire into the behavior of only one of the two parties (either the 
buyer or the seller) do not allow for the precise controlling of the 
opponents' behavior. In studies in which this control is important, 
actions and behavior of the opponents will confound the experimental 
factors . 

Clopton & Barksdale (1987) argue that negotiation studies that 
predominantly inquire into the behavior of one of the parties involved 
in the dyad may benefit substantially through the use of a microcomputer 
based methodology. A microcomputer based study (e.g., Clopton 1984; 
Schurr & Ozanne 1985) uses a format in which subjects assume the role of 
one of the two parties involved in the negotiation and negotiate with a 
microcomputer programmed as the other party. This method allows greater 
control of opponents' behavior and less confounding with relevant 
factors, thus providing the study with more power and internal validity. 
Clopton (1983) suggests that the method offers additional advantages, 
such as simplifying the administration of the experiment, quicker 
responses, availability of more time for actual negotiation, efficient 
collection of data, and accurate measurement of time-based measures 
(e.g., time elapsed between responses). Also, the subjects rate 
microcomputer simulations as more realistic and interesting . 

^Brucks (1988) advocates the use of a microcomputer based 
methodology in settings involving consumers' acquisition of information 
for similar reasons. 


In keeping with the aforementioned recommendations, the 
experimental procedure adopted in this study requires the subject to 
negotiate with a programmed microcomputer which assumes the role of the 
buyers. The computer program that is used to handle the role of the 
buyers is a contingent program. That is, it takes into account the 
offers made by the salesperson in the process of arriving at the 
counteroffers. The programming language used for the purpose is Prolog. 
Clopton & Barksdale (1987) espouse the viewpoint that an expert system 
programming language such as Prolog should be considered particularly 
suitable for the purpose of achieving both realism and control in 
contingent programs. 

The computer program written for the negotiating session is one 
that may incorporate any of several bargaining strategies and is 
equipped to negotiate on multiple terms of sale (e.g., price, delivery 
time, and so forth) . The particular bargaining strategy of the buyer 
(i.e., the computer program) that is to be employed and the choice of 
the number of dimensions for the negotiation (i.e., the separate aspects 
or terms of sale that must be settled upon) are contingent upon 
considerations of the experimental conditions and objectives. For the 
purpose of this experimental study, the only dimension utilized was 
P ri ce . In the context of testing the postulates of a study related to 
Goal Theory, a task situation comprising of a single performance 
dimension is not only adequate but may be of more value than one that 
incorporates several dimensions. This is especially true in light of 
the fact that multiple dimensions may confound the results because of 
effects which, while being relevant in the general context of 


information processing of individuals, may not be pertinent in the 
context of the conceptual model being addressed here. 

The subjects' task consists of negotiating the terms of sale for a 
fictitious office equipment on behalf of the organization that which the 
salesperson is assigned to represent. The subjects are initially 
introduced to their roles and informed about the details of the 
bargaining scenario through written instructions as well as through a 
set of preliminary messages on the screen of the microcomputer used for 
the procedures. In addition to the enactment of the role of the buyer, 
the computer program also handles the collection and storing of 
background data on the subjects as well as their responses to questions 
and probes that are used for the purpose of obtaining measures. The 
subject responds to prompts, instructions, and questions appearing on 
the screen of the computer by typing on a keyboard. 

The process of negotiation begins with an opening offer made by 
the programmed buyer, which then prompts the subject to either accept 
that offer or make a counteroffer. In the context of this particular 
example, each offer consists of a sale price (in dollars) . As a 
contingent program, the buyers' responses take into consideration the 
offers, counteroffers, and responses of the subjects. The series of 
offers and counteroffers continue until either the programmed buyer or 
the subject accepts the terms and conditions of sale (in this case, only 
the price) offered by the other. 

The contingent (i.e., interactive) nature of the computer program 
offers certain valuable advantages in the context of the bargaining 
scenario used for the experimental paradigm. In contrast with a non- 


contingent program, it is almost impossible to "outwit" the contingent 
program by repeating the same offers consistently, making either 
miniscule or ridiculously large concessions in the course of 
negotiation, or employing similar tactics that a non-contingent program 
may not be equipped to handle. Since a contingent program is also very 
realistic, the subjects tend to respond more appropriately and 
possibilities of artificial behavior on their part — a serious threat to 
rigid and mechanistic non-contingent programs — are minimized. 

A very important objective of this study comprises of measuring 
the motivational effects of certain key factors on intention to perform 
and subsequent contribution toward enhanced performance. In keeping 
with that objective, the contingent program used for the buyers' 
responses is designed to be sensitive to the expenditure of effort and 
persistence on part of the subjects in general. The different levels of 
performance that subjects may achieve in the negotiation games will 
simply reflect differential levels of effort that subjects will have put 
in rather than any significant differences in skill or training. 

The Self-Effica cy Scale for Negotiation Tasks 

Since the concept of self-efficacy is domain-specific, one needs 
to develop scales that are also domain-specific for the purpose of 
assessing self-efficacy (Bandura 1984). l Consequently, a 20-item scale 
for measuring self-efficacy in the context of negotiation activities was 

1 Refer to the section entitled "Self-Ef f icacy and Expectancy: 
Measurement Issues," in Chapter 2 for details. 


constructed (see Appendix A for a list of the items included in this 
scale) . 

The construct validity of this self-efficacy measure may be 
supported by demonstrating predictable relationships with measures that 
are already well-established. As mentioned in Chapter 2, the 
nomological net to which the self-efficacy construct belongs should 
include personality measures such as global self-esteem, as well as 
constructs that are more focussed, viz., locus of control. 

The particular instruments selected for the purpose of testing for 

construct validity include Rosenberg's (1965) 10-item Self-Esteem scale 1 

(see Appendix B for a list of items), and two sub-scales from the Sphere 

of Control scale (Paulhus 1983), viz., the "Personal Efficacy" and the 

"Interpersonal Control" sub-scales. The Sphere of Control (SOC) scale 

is both a refinement of and a development upon Rotter's (1966) scale for 

measuring "Locus of Control." According to Paulhus, 

the individual vies for control in the nonsocial environment 
in situations of personal achievement, for example solving 
crossword puzzles, building bookcases, or climbing 
mountains. Perceived control in this sphere may be termed 
personal efficacy. Second, the individual interacts with 
others in dyads and group situations, for example defending 
his or her interests at meetings, attempting to develop 
social relationships, or maintaining harmony in the family. 
In this sphere, the appropriate label seems to be 
interpe rsonal control . [1983, p. 1254] 

Silbert and Tippett (1965) report that the scale correlated 0.56 
to 0.83 with other similar measures. Rosenberg (1965) provides 
substantial evidence for the predictive validity of the scale. 


Appendix C lists the intercorrelations of the SOC scales with 
miscellaneous personality scales as reported by Paulhus (1983). 1 

The conceptual model proposed in this dissertation suggests that 
self-efficacy is determined to some extent by the global self-esteem of 
the individual in addition to the task related mastery experiences in 
the past. Therefore, the measures self-efficacy in the specific task- 
domain may be expected to be correlated to generalized self-esteem. and 
sphere of control sub-scales. 

For the purpose of assessing convergent validity, the inter- 
correlations of the self-efficacy scale with the other relevant scales 
were obtained from two different samples of subjects. Both the samples 
consisted of students enrolled in an undergraduate business course. 
Additionally, one of the two samples was comprised of subjects who took 
part in the negotiation based experiment that was adopted for testing 
the hypotheses, while the other was used as a hold out sample. The 
Cronbach's alpha measures for each of the scales and the interscale 
correlations in the experimental as well as the hold out samples are 
reported in Tables 4.1 and 4.2. 

1 The third component of Paulhus' SOC scale, "Sociopolitical 
Control," refers to the individual's perceived control in the sphere 
involving conflicts of personal goals with those of the political and 
social system. This component is not relevant in the context of the 
specific task that is of interest to this study. 


Table 4.1: Intercorrelations of the Self-Efficacy Scales with Other 
Relevant Scales in the Hold Out Sample 


1. Self-Efficacy (20 items) (.92) .45 .55 .59 .41 

2. Personal Efficacy (10 items) (.45) .44 .76 .48 
[1st sub-scale of Paulhus 1 SOC] 

3. Interpersonal control (10 items) (.75) .92 .51 
[2nd sub-scale of SOC] 

4. SOC (all 20 items) (.74) .58 

5. Self-Esteem (10 items) (.87) 

Note: 1) The scale reliabilities appear in parentheses. 

2) n=136. 

3) All correlations are significant at the 0.0001 level. 

Table 4.2: Intercorrelations of the Self-Efficacy Scales with Other 
Relevant Scales in the Experimental Sample 


1. Self-Efficacy (20 items) (.92) .24** .42 .42 .27* 

2. Personal Efficacy (10 items) (.64) .31* .76 .49 
[1st sub-scale of Paulhus' SOC] 

3. Interpersonal control (10 items) (.74) .85 .51 
[2nd sub-scale of SOC] 

4. SOC (all 20 items) (.75) .62 

5. Self-Esteem (10 items) (.84) 

Note: 1) The scale reliabilities appear in parentheses. 

2) n=112. 

3) * significant at the 0.005 level; 
** significant at the 0.05 level; 

all others are significant at the 0.0001 level. 




This study is designed to test specific experimental hypotheses 
concerning the static aspects involved in the effect of quota setting on 
motivation and performance of salespeople. 


The subjects in the experimental study were undergraduate business 
students. Since the theoretical propositions in the model presented in 
this dissertation focus on the basic motivational constructs and issues, 
this sample should exhibit the same general pattern of results as would 
be found in a sample consisting of salespeople . x 

Experimental Design 

The three independent variables that are of primary importance in 
this study are (1) the level of the externally set goal — a 5-level 
factor, manipulated between subjects; (2) the level of self-efficacy — a 
between subjects measure; and (3) external information regarding proba- 
bility of task success — a 2-level factor, manipulated within subjects. 

For the purpose of avoiding some key confounds, two more indepen- 
dent variables were considered for the design. The first of these is 
required to counterbalance the two treatment levels of the within 
subjects factor, external information. The external information 

x An appropriate follow-up for this experimental study would 
involve a separate field study utilizing a survey-based methodology. 


treatment consists of indicating to the subjects that their probabili- 
ties of succeeding are high in one of two successive tasks and low in 
the other. The counterbalancing scheme involves the introduction of the 
factor, Order (of external information treatments) that will consist of 
two levels. One level is associated with providing the subjects with 
the information that their probability of success is high in the first 
task and low in the second , the other level involves informing the 
subjects of the two kinds of probability estimates in the reverse order. 
That is, Order is manipulated between subjects and consists of two 
levels: "easy first" (or E) and "hard first" (or H) . 

The other independent variable taken into consideration is 
important in the context of assessing the stability of the self-efficacy 
measure. This factor, the time of administration of the self-efficacy 
measure, consists of two levels. For half of the subjects, the 
instrument for measuring self-efficacy was administered before the 
subjects participated in any bargaining activities; for the remaining 
half, self-efficacy was measured after the subjects had participated in 
the first of the two bargaining sessions in which they were assigned 
goals. The time of administration of the self-efficacy measure thus 
constitutes a factor that is manipulated between subjects (and consists 
of the two levels: before and after ) . 

The two treatment conditions of external information , a within 
subjects factor, was administered to all subjects (although for half of 
the subjects, it was administered in one order, and for the remaining 
half in the reverse order) . 


The dependent variables include "intention to perform, " and actual 
"performance," while the intervening measures will include "expectancy 
of task success," and "intrinsic interest." 1 

Experimental Hypotheses 

The reduced model (Figure 4.1) may be used to generate experi- 
mental hypotheses that are specific to the context of the negotiation 
scenario chosen for the experimental task. 

In accordance with the general proposition that individuals higher 

in self-efficacy respond more positively to externally assigned goals, 

the following hypothesis may be tested in the context of the task. 

HI: Salespersons who are high in self-efficacy will expend more effort 
on the negotiation task than those that are low in self-efficacy. 

The expenditure of effort may be assayed through measures of the 
total number of counteroffers made and the total amount of time spent on 
making the counteroffers by the salesperson. Since the experimental 
task was designed such that higher levels of performance reflect greater 
amounts of effort expended at the task, the actual performance levels 
achieved in the negotiations may also be considered to be additional 
measures of effort expenditure. 

The following hypothesis proposes an interactive effect of exter- 
nally set goal and self-efficacy on effort expended. In particular, it 
suggests that although externally set goals have been shown to influence 
effort expenditure in previous studies, an important aspect of this 

1 0ther auxiliary measures will include self-esteem, level of self- 
set goals, etc. 


effect is that the goal-effort relationship is stronger for those that 

are high in self -efficacy. 

H2 : As the level of the externally set goal is increased, salespeople 
who are high in self -efficacy will exhibit greater increases in 
effort expenditure than those who are low in self -efficacy. 

The theoretical model described in detail in Chapter 3 proposes an 

inverted-U shaped relationship between the independent variable (level 

of externally set goal) and relevant dependent measures, such as 

expenditure of effort and intention to expend effort. The following 

hypothesis reflects the same idea in the context of the experimental 


H3: The values of the dependent measures for effort expended (such as 

total number of rounds of negotiation, total time spent on negotia- 
tions, etc.) will reach their maximum levels corresponding to an 
intermediate goal level (i.e., corresponding to any one of the goal 
levels 2, 3, or 4 out of the total of five goal levels assigned in 
the experimental study) . 

The reduced model (see Figure 4.1) also suggests that expectancy 

of task success is not only dependent on the level of externally 

assigned goal, but is also influenced by the self-efficacy of the 

individual performing the task. In fact, the the interactive effect of 

self-efficacy and the externally assigned goal occur partially through 

the intervening variable—expectancy of task success. Therefore, it is 

important to test for the existence of the proposed relationship between 

self-efficacy of the individual at negotiation tasks in general and the 

expectancy of task success associated with a specific goal level for the 

particular negotiation task chosen for this study. 

H4 : Within each of the five levels of the externally assigned goals, 
salespeople who are high in self-efficacy will have higher levels 
of expectancy of task success than those who are low in self- 


In addition to the determinants of expectancy listed in the 

paragraph preceding hypothesis 4, expectancy of task success may also 

influenced by external information about probability of task success 

(see Figure 4.1). The rationale for this conceptualization is that 

individuals may be expected to arrive at expectancy estimates for a 

given task after considering not only the goal level assigned to them, 

but also any other external cues regarding the level of difficulty of 

the task that may be made available to them. To the extent additional 

information (regarding the level of difficulty) from sources external to 

the individual is made available to the salesperson, they are likely to 

be incorporated in the expectancy estimate that the salesperson will 

arrive at. The composite expectancy estimate thus formed (in response 

to influences from the multiple sources) will in turn affect the effort 

expended at the assigned task. The following two hypotheses summarize 

this proposition. 

H5: When salespeople are assigned the task of negotiating consecutively 
with two identical purchase representatives, but are informed that 
one of the two is a difficult negotiator (and that the other is 
easy) , their expectancies of task success will be lower correspond- 
ing to the representative that they have been made to believe is 
difficult in comparison with the expectancies corresponding to the 
other (purportedly easy) client. 

The preceding effect will remain unchanged by the order in which 
the salespeople expect to encounter the clients with the "hard" and 
"easy" labels. That is, so far as the impact of the external informa- 
tion (viz., the labelling manipulation) on expectancy is concerned, it 
will not matter whether the first client (or the second) is labelled 
"hard" (or "easy"), as long as the two clients are given one each of the 
two labels. 


H6: other conditions remaining the same, higher levels of expectancy of 
task success will contribute to increases in intention to perform 
as well as actual effort expended. 

Finally, the reduced model (Figure 4.1) identifies the intrinsic 

reward pathway, which suggests that self-efficacy moderates the level of 

intrinsic rewards that individuals will associate with a given goal 

level. Additionally, the reduced model also suggests that higher levels 

of intrinsic interest in the task enhance intention to perform and 

expenditure of effort. 

H7: For salespeople who are high in self-efficacy, higher levels of 
externally set goals will contribute to increased intrinsic 
interest in the task than those low in self-efficacy. 

H8: Intrinsic rewards will contribute positively to intention to 
perform and effort expended. 


Table 4 . 3 provides a general overview of the sequences involved in 
the experimental procedure associated with the experimental study. The 
following paragraphs explain these sequences in detail, and describe the 
manipulation of the independent factors, the procedures for obtaining 
the dependent measures, and other methodological details. 

■Step 1 ■ At the outset, all subjects are provided with a set of 
introductory messages in a brochure that describe the experimental 
setting as a part of an ongoing effort to build a sophisticated, expert 
bargaining system. Before the subjects are given any further 
information regarding the task, each subject is administered a paper and 
pencil questionnaire to collect an initial set of data including a 
demographic profile (age, sex, education, prior work experience, etc.). 
Measures of Self-esteem (see Appendix B for details), and Sphere of 


Table 4.3: Overview of the Procedural Sequences in Experimental Study 
Step # Details of the sequence 

1 - Administration of preliminary instructions and cover story. 

- Collection of initial set of data (age, sex, education, etc.). 

- Collection of paper and pencil measures of self-esteem , and 
sphere of control sub-scales. 

- For subjects in the "Before" condition of the factor Time of 
self -efficacy measurement,, the instrument for measuring self- 
efficacy is also administered. 

2 - Introductory messages presented to the subjects on the computer. 

- Participation in negotiations with the first client (Banner) . 
Collection of the following measures: Expectancy, Intrinsic 
interest, Total number of Rounds of negotiation, Total Time 
spent in negotiations, and Final Offer agreed upon. 

3 - Administration of the Goal Level treatment — one of the 5 

different target sales price is assigned to the subject for each 
of the two successive clients. 

- Administration of the External Information treatment and the 
flrder (of external information) treatment — half of the subjects 

(those in the H condition of Order) are advised that the first 
of the next two clients (AmTech) is a difficult negotiating 
opponent and that the second client (Consol) is an easy 
opponent, the other half of the subjects (in the E condition) 
are told that the first client (AmTech) is easy to negotiate 
with, and that the second (Consol) is a difficult customer. 
Collection of the following measures: Expectancy, and Intrinsic 
interest for each of the two subsequent negotiations. 

4 - Participation in negotiations with AmTech. 

- Collection of measures for the following dependent variables: 
Total number of Rounds of negotiation, Total Time spent in 
negotiations, and Final Offer agreed upon. 

5 - For subjects in the "after" condition of the factor Time of 

self-efficacv measurement the instrument for measuring self- 
efficacy is administered. Those in the "before" condition skip 
this step. 

6 - Before participating in negotiations with the final client 

(Consol), some key intervening variables in the context of that 
task are measured once again to assess the effect, if any, of 
the first negotiation. 

7 - Participation in negotiations with the last client (Consol) . 

Collection of measures for the following dependent variables: 
Total number of Rounds of negotiation, Total Time spent in 
negotiations, and Final Offer agreed upon. 

8 - Debriefing of the subjects. 


Control (Paulhus 1983) are also collected. For half of the subjects, 
those in the "before" condition of the factor Time of measurement of 
self-efficacy, the instrument for measuring self-efficacy, consisting of 
20 items (see Appendix A) , is administered at this point using a 7-point 
Likert format scale. 

Next, the subjects are provided with information regarding the 
role of the salesperson that they are expected to assume, the details of 
the organization that they are supposed to represent, and the product 
whose terms of sale they are assigned to negotiate. In particular, the 
subjects are informed that they will be required to negotiate and settle 
on the selling price of the product with each of three different 
purchasing representatives from separate organizations. They are also 
informed that their negotiations will be conducted via the computer 
terminal. They will receive offers and messages from their clients on 
the computer screen, while they in turn may transmit their messages and 
counteroffers by typing on the computer's keyboard. 

Each subject is informed that the scenario involved in negotiating 
via the computer terminal requires the subject to assume the role of a 
salesperson employed by IMC — a large company engaged in the business of 
manufacturing and selling document processing equipment. The particular 
product whose price the subject (salesperson) would have to negotiate is 
known as a "Jaeger." The manufacturing cost of a Jaeger is $15,000. It 
represents a product that incorporates a number of technological break- 
throughs and the installation of which requires requires a high degree 
of customization. Because of the recency of the introduction of the 
product and also because the product's configuration is so different 


from those of equipment used thus far, the Jaeger does not have any 
standard or reference price yet. IMC has chosen to rely on its sales 
department to develop a pricing scheme on a case by case basis, that 
would enable sales managers and their sales teams to obtain as much 
revenue as possible. In the document processing field, it is quite 
common to price a product which incorporates new technology at a much 
higher level than its manufacturing cost. 

IMC's clients have different characteristics, but most of them 
have some amount of interest in upgrading the document processing 
facilities that they currently use in their offices. The purchase 
representatives of the clients may be expected to possess different 
bargaining styles and have unequal amounts of negotiating experience. 
The subject is advised of the fact that he/she could be made aware of 
matters concerning a particular client's bargaining style before negoti- 
ations with that client begins, provided such information is available 
to the subject's superiors at IMC. All clients have been thoroughly 
informed about the product's characteristics through different kinds of 
marketing communication efforts undertaken earlier. At this time, a 
settlement of the product's price is the only critical aspect of the 
selling process that has not been completed. This settlement will be 
done through negotiations between IMC's salespersons and the clients' 
purchase representatives . 

Before the subjects proceed to the next stage of the experimental 
task, each subject is given an outline of the sequence of events which 
he/she may expect to encounter when the interactions with the computer 
begins. Also, instructions regarding the details of carrying on negoti- 


ations via the computer terminal are also given to the subjects before 
the start of the actual negotiations. 

Initially, the subject will be greeted with a welcome message on 
the computer screen and be prompted to enter his/her identification 
number (provided by the experimenter) on the keyboard. The computer 
program will respond with a number of message screens that reiterate 
some of the key aspects of the negotiating scenario. Next, the computer 
screen will bring a message from the salesperson's immediate superior — a 
regional sales manager at IMC. The sales manager will assign a specific 
client and a particular sales quota for that client to the subject. 
When the sales manager's instructions have been given to the subject, 
the individual subject may expect to receive a message from the first 
client so that negotiations on the price of the product may commence. 
After the completion of the negotiation session with the client, the 
subject will once again hear from the sales manager for instructions 
regarding subsequent tasks. 

The salesperson's interactions with the client's purchase repre- 
sentative will begin with a message from the representative appearing on 
the computer's screen. The purchase representative will introduce 
himself/herself and then begin the negotiation by making an opening 
offer (in terms of a dollar figure) for the product. The subject will 
have the option of either accepting that offer or making a counteroffer 
or, as a third alternative, typing a message on the computer's keyboard 
in response to the client's message and offer. This message may be 
typed by the subject at the time he/she makes counteroffer. 


Under normal circumstances, the subject (enacting the role of a 
salesperson) would be expected to make a counteroffer for the selling 
price at a level that is much higher than the client's initial offer. 
In response to the subject's offer, the client may be expected to 
transmit a message on the subject's computer screen in order to explain 
whether or not he/she agrees to that counteroffer. In case the purchase 
representative decides to reject the initial counteroffer, as is most 
likely in the initial stages of the negotiation process, he will make a 
revised (possibly higher) offer for the amount at which they would like 
to buy the product. However, if the subject makes an extraordinarily 
low initial counteroffer, the purchase representative of the client may 
be willing to accept it as well. Ordinarily, both the parties involved 
in the negotiation are likely to trade offers and counteroffers a number 
of times before any one of the two negotiators eventually decide to 
accept the offer made by the other. 

The subject is informed that the messages which he/she has the 
option of typing for the client to read do nci actually reach the 
purchase representative while the negotiations are in progress. On 
account of technical limitations, the subject's comments and messages 
are recorded in a "message center"--a storage device which the purchase 
representative may play back after the negotiation session. At that 
time, the purchase representative would be able to read the entire 
script consisting of the dialog between the two negotiating parties in 
the sequence in which it had occurred. Also, the sales manager at IMC 
is empowered to obtain a copy of a recorded session at his/her discre- 
tion in order to evaluate any particular salesperson. Although the 


messages do not reach the client while the negotiations are in progress, 
the subject is urged to record a message or enter a comment whenever 
he/she considers one appropriate so that either of the two negotiating 
parties may have the opportunity of examining and learning from a given 
negotiation experience after it is over. 

This scenario concerning the unavailability of the on-line 
messages by the purchase representative is introduced on account of the 
fact that the computer program (which handles the interactions with the 
subject) is incapable of interpreting and responding to verbal messages. 
The program, however, is capable of reading the counteroffers made by 
the subject in numerical terms and can even adapt itself in response to 
the pattern of concessions being offered by the subject. Since the 
program is not interactive in so far as verbal messages are concerned, 
this scenario is introduced so that interactions with the subjects at a 
semantic level may be avoided. The messages that are transmitted by the 
programmed client are actually chosen by the computer program from a 
message bank corresponding to the particular client involved in 
negotiations at any time. Each client will have his/her own message 
bank which effectively constitutes the range of vocabulary that the 
programmed client possesses. The program's decision regarding the 
choice of the message to be transmitted by the client is based on the 
computation of the client's counteroffer and concession pattern and is 
meant to be consistent with its own actions. That is, the messages 


transmitted by the purchase representative are generally meant to be 
comments that accompany its own offers. 1 

Thus far, all the instructions that are imparted to the subjects, 
the collection of the measures, etc., are done using a paper and pencil 
format. All the instructions and questionnaires, etc., are presented in 
one brochure. 

£t_ep_2. After the completion of the preceding step, the subjects 
are instructed (at the end of the brochure given to them in Step 1) to 
read the opening message on the computer screen. At this point, the 
subjects may read the opening message, type in the identification 
numbers provided by the experimenter, follow the instructions that 
appear on the computer screen, press the appropriate keys on the 
keyboard, and be guided through the rest of the experiment by the inter- 
active computer program. 

In the opening stages of the subject's interaction with the 
computer program, the subjects are first led through a series of 
messages that summarize the important aspects of the process of negotia- 
tion, and the details of the role assigned to them are reiterated. 

Next the subject is informed that their subsequent instructions 
will come from the sales manager, i.e., the immediate superior of the 
salesperson whose role the subject has assumed. 

The messages from the sales manager are associated with a 
different set of colors on the computer's screen from what had been used 

Refer to Appendix G for a sample of the types of messages 
included in a typical client's message bank. 


thus far. Each time the identity of the individual (interacting with 
the subject) changes, the colors and other screen configurations change 
accordingly. However, the same individual is consistently associated 
with the same configuration whether or not that person interacts with 
the subject at one stretch or sends messages to him from time to time. 

After a brief set of introductory messages, the sales manager 
informs the subject that purchase representative with whom the sales- 
person will be negotiating shortly is from an organization known as 
Banner Industries. The sales manager also assigns the subject a sales 
quota equal to $18,000. The quota assigned for the first negotiation 
session with Banner is the same for all subjects. 

The negotiation session with Banner is intended to serve two 
purposes. First, it provides a vehicle for the subjects to become 
familiar with the use of the computer terminal for the purpose of nego- 
tiating. Second, the uniform target performance level assigned to all 
subjects is intended to serve as a baseline or frame of reference, in 
the context of which the goals to be assigned for the subsequent 
negotiations may be perceived and evaluated. In other words, five 
different groups of subjects will be assigned different goals for the 
subsequent negotiations, all of which will be greater than the baseline 
goal (the $18,000 quota assigned for Banner), but the amounts by which 
they will exceed the baseline goal will vary. Thus, it will be the 
differing amounts of increases in the target performance levels that 
will constitute the five treatment levels of the factor Externally Set- 
Goal Level . 


At the end of his instructions, the sales manager announces that 
he will be transferring his end of the computer terminal briefly to the 
personnel of the Human Resources Department of IMC (the salesperson's 
and the sales manager's own organization). The Human Resources (or HR) 
department routinely administers a short, confidential questionnaire to 
all new employees for their first few job assignments. After the 
interactions with the HR department have been completed, the subject may 
except to be connected directly to the purchase representative from 

The colors of the computer's screen change once again to a new 
configuration that uniquely identifies the HR department. Next, the HR 
department will introduce itself and reiterate the point that the confi- 
dentiality of the responses to the subsequent questions will be strictly 
maintained and not even the sales manager will have access to those 
responses. The survey is administered merely for the purpose of 
assessing the characteristics of the assigned task and not the task 
performer. In actuality, this scheme is introduced for the purpose of 
measuring the following constructs: subjective difficulty of goal, 
expectancy of task success, likelihood of achieving the quota, the 
confidence rating associated with the expectancy and likelihood 
estimates, intrinsic interest in the task for the quota assigned, and 
intention to expend effort (see Appendix D for further information 
regarding the instruments used for measuring the preceding constructs) . 
After the measures have been collected, the HR department informs the 
subject that they may next expect to hear from Banner. 


The bargaining session with the purchase representative from 
Banner begins with introductory messages and an opening offer from the 
client. Once again, the session with Banner employs colors and configu- 
rations for the computer's screen that uniquely identify this particular 
client. The subject has the option to accept Banner's offer or make a 
counteroffer. The Banner representative evaluates the subject's 
counteroffer and makes a revised offer that reflects a small but 
positive concession on part of Banner. The exchange of offers and 
counteroffers continue until either the computer program representing 
the Banner purchasing agent or the subject accepts the opponent's offer. 
(Appendix E describes the algorithm used for the bargaining procedure in 
detail) . 

The dependent measures collected by the computer program in the 
course of the negotiation sessions include the total number of offers 
made by the subjects, the total amount of time spent in the bargaining 
session, the total number of characters typed by the subject (in the 
form of messages directed to the client) , and the actual sale price at 
which the negotiation concludes. 

Step 3 . At the point of conclusion of the bargaining session 
with the representative from Banner, the subject is informed that the 
subsequent instructions will originate from the sales manager. The 
colors on the screen of the computer are switched back to those 
associated with the manager at this point. 

The sales manager's message congratulates the subject for having 
completed the preceding negotiation and informs him/her that the next 
task will consist of similar negotiations, first with the purchasing 


representative from Amalgamated Technologies (AmTech) , and thereafter 
with a purchase representative from Consolidated Machineries (Consol) . 

At this stage of the experiment, the externally set goal level 
treatment is administered. The subject is randomly assigned to one of 
the five different goal level conditions selected for the experiment. 
In particular, the subject is advised by the sales manager that he/she 
is being assigned a specific sales quota for each of the two subsequent 
negotiations. That is, the assigned quota is the selling price which 
the subject is expected to realize (i.e., settle on after negotiation) 
for each of the two clients. The actual dollar values of the sales 
quota assigned for the different levels of the goal level treatment are 
listed in Table 4.4. 

Table 4.4: Dollar Values of Sales Quotas Corresponding to Different 

Levels of the Factor Externally Set Goals 

Level of Goal Dollar Value of Sales Quota 

1 $20,000 

2 $25,000 

3 $35,000 

4 $55,000 

5 $75,000 

The manipulation External Information and Order (of external 
information) are also simultaneously administered at this point. 
External information is a within subject factor and its manipulation 


consists of imparting to the subjects either the information that AmTech 
(the first of the next two clients) is a hard bargainer and Consol (the 
last client) is easy, or, alternatively, the information that AmTech is 
the easy bargainer and that Consol is hard. In effect, these two sub- 
groups of subjects comprise the two levels of the between subject 
factor, Order (of external information) . 

The subjects in the "Hard first" (or H) condition of the factor 
Order (of external information) are informed that AmTech, the first 
client (of the two subsequent negotiation trials) and is a hard 
bargainer, and that Consol, the final client, is an easy buyer to 
negotiate with. Those in the "Easy first" (or E) condition of the 
factor Order are told that the first client, AmTech, is the easier 
client and that Consol, the final client, is the tough bargainer. 

In reality, the computer program that represents one does not 
differ from the other except in superficial details, such as style of 
presentation of offers, the colors (and graphics) on the screen, etc. 
The actual algorithms that are used for the negotiations are exactly 
identical (see Appendix E for details of the algorithm) . This scheme 
permits the manipulation of the within subjects factor External 
Information as well as the counterbalancing of the treatment levels via 
the use of the factor, Order . 

Next, the subject is once again ostensibly linked via the computer 
terminal with personnel from the HR department who are purportedly 
collecting information for evaluating the characteristics of the task. 
The questionnaire administered to the subject is used for the purpose of 
measuring the following intervening constructs in the context of the 


negotiation sessions with AmTech and Consol: subjective difficulty of 
goal, expectancy of task success, likelihood of achieving the quota, the 
confidence rating associated with the expectancy and likelihood 
estimates, intrinsic interest in the task for the quota assigned, and 
intention to expend effort. The instruments used for collecting these 
measures are similar to the ones used for measuring the same variables 
corresponding to the negotiation with Banner. Appendix D lists the 
scales and items that were used for the collection of the measures in 
the context of the negotiation with AmTech. The procedure for 
collecting measures corresponding to the negotiation with Consol were 
identical to the one used for AmTech. 

Step 4 . This step consists of having the subjects negotiate with 
the computer program that now represents the purchasing agent from 
AmTech. This task generates the following dependent measures that are 
also recorded by the computer program. 

Expenditure of effort is assayed by observing the total number of 
counteroffers made. Indications of persistence (expenditure of effort 
and cognitive resources) are also obtained from the examination of 
response times of the subjects. The total time spent per negotiation 
and the number of counteroffers made are considered to be convergent 
measures of expenditure of effort. The initial offer (i.e., the 
starting point of the negotiation) is also a reflection of the level of 
performance that the subject intends to achieve. As an auxiliary 
measure of the intention to perform the subjects' internal (i.e., self- 
set) goal is also measured. Performance is measured by actual profit 
levels achieved in the process of arriving at a settlement. 


Step 5 . For subjects in the "after" condition of the factor, Time 
of measurement of self -efficacy , the instrument for measuring self- 
efficacy is administered. 1 Since this factor is orthogonal to the other 
independent variables that were manipulated, it can be of aid in 
analyzing whether the self-efficacies of subjects change on account of 
the negotiation experiences they encounter in the preceding steps. In 
particular, if this factor — Time of measurement of self-efficacy — does 
no_t affect self-efficacies measured across the sample, one may conclude 
that the short-run bargaining experiences provided in the preceding 
steps do not alter self-ef f icacies of subjects. That would be in 
accordance with the concept of the construct which is envisaged to be 
relatively enduring in nature, at least so far as short term experiences 
are concerned. The argument may be further strengthened if the 
interaction of the factors Time of measuring self-efficacy and External 
Information do not affect self-efficacies of subjects across the sample. 

■Step 6. This step merely consists of measuring one of the key 
intervening variables viz., expectancy of task success, in the context 
of the negotiation with Consol, the final client. The motivation for 
this is similar to the one in the preceding step. The difference 
between the expectancy measures (with regard to the negotiations with 
Consol) that are collected before and after the experience of bargaining 
with AmTech may provide information about the relative stabilities of 
the constructs Expectancy and Self-ef ficacy. 

1 Those in the "before" condition had their self-efficacies 
measured in Step 1 and consequently they skip this step. 


Step 7. Finally, the subject negotiates with the computer program 
that represents the purchasing agent from Consol. As mentioned earlier, 
the actual algorithm used for negotiating by the computer program is the 
same as the one used for AmTech. The only differences are those that 
involve superficial aspects such as presentation styles, etc. The 
relevant dependent measures in the context of this final negotiation 
session are collected. 

Step 8 • The subjects are finally asked to rate the entire 
experiment on realism, interest and are subsequently debriefed. 


Several theoretical concerns dictate the use of an experimental 
setting for testing specific hypotheses generated from the conceptual 
model. A performance task involving a bargaining scenario, in which the 
subjects negotiate the terms of sale with buyers (represented by a 
programmed microcomputer), is considered appropriate for the experiment. 

Independent variables consist of level of externally set goal , 
(i.e., target performance levels in terms of negotiation outcomes); 
self-ef ficacy (a between subjects measure) ; and external information 
(involving assessments of the participant's probability of task 
success) . Dependent measures include actual performance levels (the 
final sale price agreed upon through negotiations) , intention to 
perform, and actual effort expended. Expectancy of task success and 
intrinsic rewards are included among variables that comprise of the 
intervening measures. 


In this chapter, the results of data analyses are reported and the 
implications of the observed results are discussed. The first of the 
three main sections of this chapter focuses on the self-efficacy 
measure. In particular, this section attempts to provide evidence in 
support of the validity and stability of the self-efficacy measure in 
addition to what was provided in the preceding chapter. The second 
section of this chapter is a detailed report of important results from 
analyses of data undertaken for the purpose of testing the experimental 
hypotheses . 

Analyses of the Self-Efficacy Measure 

Stability of Self-Efficacy 

According to the theoretical model outlined in Chapter 3, self- 
efficacy for negotiation tasks is a relatively enduring measure that 
should not change in response to the subject being presented with a few 
tasks with varying levels of goal difficulty. That is, self -efficacies 
of individuals are considered to remain stable over the short run. 
Expectancy measures, on the other hand, are expected to change every 
time the goal level associated with the task changes, provided all other 



factors are held constant. In fact, this is considered to be one of the 
major distinctions between the self-efficacy and expectancy measures. 

In the experimental study described in Chapter 4, self -efficacies 
of participants were measured with a scale that was constructed speci- 
fically for that purpose. Expectancy of task success was measured by 
having the subjects report their subjective probabilities of being able 
to achieve the assigned goals. The subjects were asked to indicate a 
number from (zero) through 100 (one hundred) corresponding to their 
probability estimates (in percentage terms) of their being able to reach 
the assigned target performance level. As a convergent measure of 
expectancy, the subjects also rated the likelihood of task success on a 
7-point Likert format scale, with the phrases "extremely likely" and 
"extremely unlikely" anchoring its two ends. 

As mentioned in Chapter 4, the self-ef f icacy scale used on the 
experimental sample was found to have a Cronbach's alpha reliability 
measure of . 92 . To ascertain whether the self-efficacy measure was 
sensitive to the treatment conditions of external information (as well 
as of the level of assigned goal) , the factor time of measurement of the 
self-e fficacy measure was introduced. Those in the before condition of 
this factor had their self-efficacies measured before they were 
administered the goal level treatment (i.e., before any negotiation 
tasks) . Those in the after condition of this factor had their self- 
efficacies measured after the first of the two bargaining sessions which 
consisted of the assigned goals (i.e., after the second negotiation 
session or the one in which the client was AmTech) . Also, this factor 
was orthogonal to the other factors that were manipulated and the 


assignment of subjects to the two treatment conditions of this factor 
was random. An ANOVA analyses with self-ef f icacy as the dependent 
variable and time of measurement of self-efficacy as the predictor 
variable should therefore yield non-significant results. A significant 
difference in self-efficacies across conditions of the factor time of 
measurement of self-efficacy would not be supportive of the proposition 
that the self-efficacies of the subjects had remained stable during the 
course of the experiment. 

In order to make a stronger statement about the stability of the 
self-efficacy measure, however, the ANOVA model should also include the 
factors Goal level and Order of external information and their 
interactions with the time of measurement of self-efficacy factor among 
the independent variables. This is necessary to demonstrate that there 
does not exist differential attenuations of self-efficacies when the 
self-efficacies of subjects in high goal level conditions are compared 
between before and after conditions of time of measurement on one hand, 
versus when the same self-efficacies are compared for subjects in low 
goal level conditions on the other. Such a differential would involve a 
significant effect of the interaction of goal level and time of 
measurement of self-efficacy on self-efficacy in the ANOVA model. Also, 
since time of measurement of self-efficacy is orthogonal to the factor 
Order of external information, half of the subjects in the after 
condition of time of measurement of self-efficacy had their self- 
efficacies measured after the purportedly hard bargainer, while the 
remaining half of the subjects in the after condition had self- 
efficacies measured after what they were made to believe was the easy 


negotiation. Thus, there may exist a possibility of differential 
attenuation of self-efficacies if we consider these two groups as well. 
In this case, the interactive effect of time of measurement of self- 
efficacy and order on self-efficacy in the ANOVA model has to 
significant for that to be true. 

The ANOVA model consisting of Self-Efficacy as the dependent 
variable and each of the three factors — Time of measurement of self- 
efficacy, Goal level, and Order, along with the three 2-way interactions 
of these three independent variables was run. Neither the model nor any 
of the individual type III sums of squares were significant. Also, 
Bonferroni T-tests for the variable self-efficacy, done separately for 
each of the three independent variables, did not yield any significant 
difference across any pair of treatment conditions of any of the three 
factors. Consequently, the self-efficacy measure may be considered to 
be unconfounded with other manipulated factors and may be regarded as a 
stable, trait-like measure, at least so far as this experimental study 
is concerned. Additionally, the factor Time of measurement of self- 
efficacy may be eliminated from considerations in subsequent analyses. 

Effects of Self-Efficacy 

In the first of the three negotiation sessions (in which the 
subjects negotiated the selling price with Banner) , all subjects were 
assigned the same goal or target performance level as well as the same 
bargaining task situation. In the context of the first negotiation 
session, results of data analyses indicate that subjects who were high 


in self-efficacy were significantly different (from those who were low 
in self-efficacy) in terms of the following characteristics: 1 

1) they reported higher expectancies of task success , when expectancy 
was measured as a subjective probability (EXPEC-PROB) or as a likeli- 
hood estimate (EXPEC-LIKE) ; 

2) they reported being more confident (CONFID) with regard to their own 
probability and likelihood estimates; 

3) they rated the performance goals (i.e., the assigned sales quotas) as 
less difficult (DIFFICULT) ; 

4) they spent more time in the negotiation process (TOTAL TIME), 
participated in more rounds of bargaining (ROUNDS) , and concluded 
their negotiations at higher prices (LAST OFFER) . 

Self-efficacy, however, failed to predict intrinsic interest in 
the context of achieving the assigned quota levels (INTRINSIC) as well 
as the intention to expend effort (INTENT) at the 0.05 level of signi- 

The preceding results were obtained from GLM analyses of the 
experimental data with self-efficacy as the predictor variable and each 
of the aforementioned criterion variables in separate models. Table 5.1 
summarizes the results of these analyses and includes the parameter 
estimates of the the predictor variable self-efficacy from each of the 
models . 

The terms in upper case letters and enclosed within parentheses 
denote the corresponding dependent variables as indicated in the 
summarized report in Table 5.1. 


Table 5.1: Effect of Self-Efficacy on Other Variables 

Criterion Parameter estimate T for HO: Significance Std. error 
variable of Self -Efficacy Parameter=0 Level of paramete r 














































Tests of Hypotheses 

The hypotheses have been tested within the context of the two 
negotiation sessions in which goals were assigned. In other words, the 
dependent measures from the second and the third negotiation sessions 
(corresponding to the clients, AmTech and Consol, respectively) have 
been analyzed in order to test the hypotheses. 

Performance Levels 

The indicators for performance that were chosen include ROUNDS 
(the total number of counteroffers made by the subject in response to 
the offers made by the computer) , LAST OFFER (the price at which 


settlement was reached) , and TOTAL TIME (the total amount of time spent 
by the subject in considering the counteroffers) . 

Table 5.2 lists summarized results for the performance measures 
(for both negotiation sessions) for the ANCOVA model of the following 
form: Dependent measure = G S G*S, 

where, G = Externally assigned goal (5-level treatment) ; 
S = Self-efficacy measure (a covariate) ; and 
G*S = the interaction term. 

An examination of the Type III sums of squares of ROUNDS, and LAST 
OFFER reveal that for both sessions, the interaction term and the main 
effect of S (self-efficacy) are significant. However, for the dependent 
variable TOTAL TIME, only the main effect of S was significant for one 
of the two sessions. 

Table 5.2 evidently furnishes strong support for hypothesis 1 
which proposes that salespersons higher in self-efficacy will expend 
more effort than those who are low in self-efficacy. 

ROUNDS . For the purpose of illustrating the interactive nature of 
the effect of self -efficacy on the several dependent variables, the 
entire sample was divided into four groups corresponding to the lowest 
through the highest quartile scores on self-efficacy. Appendix F 
reports some of the relevant statistics associated with the self- 
efficacy scores of each quartile. 

Tables 5.3 and 5.4 report the mean values of ROUNDS for the two 
negotiation sessions respectively for the different levels of the factor 
Goal Level crossed with the four self-efficacy blocks. 


Table 5.2: Results from Analyses of the First ANCOVA Model for the 

Performance Measures 









Sia. level 


Sig, level 



















R 2 - 0.47 
























R 2 >■ 0.38 
























R 2 » 0.35 
























R 2 = 0.32 
























R 2 = 0.37 
























R 2 = 0.27 







Table 5.3: Mean Values of ROUNDS for the Negotiation with AmTech 

(2nd Session) 

Goal Levels 






Block 1 







Block 2 








Block 3 






Block 4 






Row means 






9.59 16.00 15.26 28.78 24.30 18.75 

Table 5.4: Mean Values of ROUNDS for the Negotiation with Consol 

(3rd Session) 

Goal Levels 






Block 1 








Block 2 







Block 3 






Block 4 







Row means 





8.45 11.91 14.87 26.26 28.00 17.76 


Examination of the mean values of ROUNDS in Tables 5.3 and 5.4 
clearly reveal that those subjects who are high in self-efficacy tend to 
make more counteroffers. Also, in Table 5.3, the goal level 4 
corresponds to the highest performance of subjects with regard to the 
number of counteroffers made, and the same pattern of results is evident 
for each of the four self-efficacy blocks. In Table 5.4, the goal 
level 5 corresponds to the highest value. However, the mean number of 
rounds for goal level 5 is only slightly higher than that for goal level 
4. Additionally, for the self-efficacy blocks 1 and 3, goal level 4 
corresponds to the highest values of the number of rounds of counter- 
offers . 

In accordance with hypothesis 2, blocks corresponding to higher 
self-efficacy scores clearly exhibit greater increases in ROUNDS as the 
level of goal difficulty increases (i.e, as the goal level rises from 1 
through 5) . 

LAST OFFER and TOTAL TIME . Tables 5.5 and 5.6 report the mean 
values of LAST OFFER from the two relevant negotiation sessions for the 
five levels of the factor goal level crossed with the four levels of the 
blocking factor generated from the self-efficacy scores; while Tables 
5.7 and 5.8 list mean values for the dependent measure TOTAL TIME for 
the two negotiations. 

The dependent measure LAST OFFER exhibits the same general pattern 
of results as ROUNDS. As before, the goal level 4 generated the highest 
performance from the subjects in both negotiation sessions, for each 
self-efficacy block as well as the entire sample (Tables 5.5 and 5.6). 


Table 5.5: Mean Values of LAST OFFER for the Negotiation with AmTech 

(2nd Session) 

Goal Levels 






Block 1 








Block 2 







Block 3 






Block 4 






Row means 





Means 20684 21629 22074 32330 24508 24294 

Table 5.6: Mean Values of LAST OFFER for the Negotiation with Consol 

(3rd Session) 

Goal Levels 






Block 1 








Block 2 







Block 3 






Block 4 







Means 22544 23544 24526 33770 27529 

Row means 







Table 5.7: Mean Values of TOTAL TIME for the Negotiation with AmTech 

(2nd Session) 

Goal Levels 






Block 1 







Block 2 







Block 3 






Block 4 






Means 79.55 164.23 123.78 279.40 213.73 

Row means 






Table 5.8: Mean Values of TOTAL TIME for the Negotiation with Consol 

(3rd Session) 

Goal Levels 






Block 1 


127. 07 





Block 2 







Block 3 






Block 4 






Means 64.65 110.16 106.89 226.30 194.23 

Row means 







The pattern of results for the dependent variable TOTAL TIME, 
however, is weaker. Firstly, the goal level x self-efficacy interaction 
for the two negotiation sessions were not significant at the 0.05 level 
(Table 5.2) . Also, the mean value of total time does not increase 
progressively as the goal level is increased from 1 through 4. The 
order of goal levels that generated the highest through the lowest 
levels of the dependent variable TOTAL TIME (for both sessions) is 4, 5, 
2, 3, 1 (Tables 5.7 and 5.8). 

All the preceding dependent variables provide some evidence in 
support of hypothesis 3 which posits inverted-U shaped curvilinear 
relationships between goal level and measures of performance, with 
performance peaking at intermediate goal levels. Further analyses and 
comments pertaining to this issue are reported at the end of this 

For the purpose of further analyzing the effects of G and S on the 
performance measures, the dependent measures from the previous ANCOVA 
analyses were analyzed in the context of a second model of the following 
form: Dependent variable = G S (G) ; 

where the symbols representing the independent variables have the same 
meanings as in the preceding analyses. 

The parameter estimates from this model provide information about 
whether or not the slope of the relationship between self-efficacy and 
the dependent variable is significantly different from zero within each 
level of goal. Table 5.9 reports the relevant parameter estimates. The 
results indicate that for lower goal levels, the relationship between 
performance and self-ef f icacy is not significantly different from zero. 


For higher goal levels, however, higher levels of self -efficacy contri- 
bute to improved performance. This is consistent with the proposition 
of the first hypothesis. 

Table 5.9: Effect of Self-Efficacy on Performance Measures Within Each 
Goal Level - Parameter Estimates from the Second ANCOVA Model 




2nd 3rd 


2nd 3rd 

2nd 3rd 

S (G 1) 

3.7 1.72 

29.17 -468.51 

11.08 3.47 

S (G 2) 

4.08 -0.42 

772.70 -810.47 

30.37 12.50 

S (G 3) 

-0.11 -0.31 

152.83 828.89 

-6.31 -12.93 

S (G 4) 

8.85** 7.49* 

3578.53* 5429.00* 

59.03* 33.36 

S (G 5) 

8.99** 9.56** 

6276.08** 7926.16** 

51.66* 44.29* 

Note: The parameters S (G 1), S (G 2) , etc., refer to self-ef f icacy 
within goal level 1, self-ef f icacy within goal level 2, etc., 

* - Significant at the 0.05 level; 

** - Significant at the 0.0005 level. 

Expectancy of Task Success 

Hypothesis 4 proposes that within a given level of goal diffi- 
culty, expectancy of task success will be positively correlated with 
self-efficacy. That is, subjects with higher self-efficacies will 
report higher expectancies for the same level of goal difficulty. 


To test this hypothesis, a GLM analysis with Expectancy as the 
dependent variable and Self-Efficacy as the independent variable was 
done separately for each goal level. The results are summarized in 
Table 5.10. 

Table 5.10: Effect of Self-Efficacy on Expectancy Within Goal Levels 


(2nd Negotiation) 

F-Value Sig. level 

(3rd Negotiation) 

F-Value Sig. level 





















Analyses of the ANCOVA model: Expectancy = G S (G) also confirmed 
the above results, viz., in seven out of the ten cases, the slope of the 
relationship between self-efficacy and expectancy was significantly 
positive. Thus there seems to be a moderate level of support for the 
proposition that higher self-efficacies contribute to higher 
expectancies within goal levels. The fact that the expectancy measures 
from the third session provide stronger support for this hypothesis may 
be an indication of the possibility that subjects' estimates of the 
probability of task success improve or stabilize (i.e., have lower error 
variances) as the subjects become more and more familiar with the task 


Hypothesis 5 suggests that external information regarding the 
probability of task success (i.e., estimates of probability of task 
success provided by the experimenter ) should also affect the subjects' 
expectancies. The within subjects factor, EXTINFO, involves informing 
the subjects that one the two clients (of the second and third sessions) 
was a hard negotiator, while the other was easy. In actuality, the same 
task was repeated for the two sessions. Hence, a difference between for 
the expectancies for the second and the third negotiations sessions 
should confirm this hypothesis. 

The ORDER manipulation, however, involved informing half the 
subjects that AmTech (2nd client) was the easy negotiator and that 
Consol (3rd client) was the difficult one; while the remaining half was 
informed that AmTech was hard and that Consol was easy. Thus, if 
EXTINFO represents the within subjects factor (repeated measures of 
expectancy for the 2nd and the 3rd task, or EXPEC2 and EXPEC3 
respectively), then the repeated measures GLM model: 

EXPEC2 EXPEC3 = ORDER (Repeated = EXTINFO) is the appropriate model for 
the analyses. 

Expectedly, the main effect for EXTINFO (in the corresponding 
MANOVA tests) was not significant. Also, the EXTINFO x ORDER effect was 
significant at the 0.0001 level. The analysis of the contrast effect 
between EXPEC2 and EXPEC3 also confirmed the preceding result. 

Another way of approaching the aforementioned analysis is by 
looking for a main effect of EXTINFO when the repeated measures GLM 
model is run within each of the two levels of ORDER. Once again, 


EXTINFO had a significant effect on the expectancy measures within each 
level of ORDER. 

The next analysis incorporated a somewhat more involved model with 
self-efficacy, goal level, and their interaction as independent 
variables (and EXTINFO as the within subjects independent variable) and 
was done separately for each level of ORDER. It was found that the 
EXTINFO x SELF-EFFICACY interaction was significant. Although this 
effect had not been hypothesized in the context of the experimental 
study, this finding is consistent with the conceptual model discussed in 
Chapter 3. Additionally, this result has intuitive appeal — those who 
are low in self-efficacy are likely to revise their expectancies 
downward to a greater extent (than those who are high in self-efficacy) , 
when cautioned about the high level of difficulty of an impending task. 

In sum, the results of the analyses provided strong support for 
hypothesis 5. 

Intent ion to Perform and Intrinsic Rewards 

In general, hypothesis 6, 7, and 8 received partial support from 
the results of the analyses of data. Expectancy of task success was 
found to be significantly related to the indicators of performance, 
mostly for goal levels 4 and 5. For lower levels of goals, the 
parameter estimates of expectancy within goal levels did not approach 
significance. Also, the 'Expectancy-Intention to Perform' relationships 
within goal levels approached significance in about half of the cases, 
while being actually significant at the 0.05 level in the other half. 


The preceding conclusions were obtained from models of the form: 
Dependent variable = Expectancy (Goal level) . 

With goal level as an additional independent variable in the 
preceding model, the significance of the parameter estimates are 
considerably weakened. A possible reason for that occurrence is the 
strong multicollinearity between goal level and expectancy. 

Models with INTRINSIC (the level of intrinsic interest in the 
level of the assigned goal) as the dependent variable, and self-ef f icacy 
and goal level as independent variables did not even approach signifi- 
cance. Consequently, hypothesis 7 remained unsupported. 

When INTRINSIC was used as the independent variable, it performed 
poorly in predicting the performance measures for the first negotiation 
session. However, for the second session, two of the three performance 
measures were predicted significantly, while in the third case the 
relationship merely approached significance. 

In the context of the aforementioned weak results, it was somewhat 
surprising to observe that INTENT (intention to expend effort) was 
predicted strongly by INTRINSIC. 

One of the commonalities of the results from the three negotiation 
sessions include the lack of support for predicted relationships between 
INTENT and the primary independent variables, Goal Level and Self- 
Efficacy. Similarly, INTRINSIC seemed to exhibit a lack of correlation 
with the two primary predictors. It is speculated that the measures 
failed to capture the essence of the constructs they were meant to 
represent — a possibility that is further explored in the following 


INTENT was measured by having subjects respond to an item that 
asked them how hard they would work to realize the assigned goal. It is 
possible that the subjects (or at least a significantly large number of 
them) interpreted the question as: Given the level of goal r how hard 
would you work to realize it? Thus, although task goals of different 
difficulty levels would require differential amounts of effort, in each 
case the subjects could indicate "extremely hard" and yet be semanti- 
cally correct. 

Similarly, INTENT was measured on an instrument consisting of 
three items which essentially asked subjects to indicate the level of 
enjoyment they would associate with the goal level. As in the case of 
INTRINSIC, there is the possibility that the subjects considered the 
questions separately for each goal level and not across different task 
goals. This speculation finds some justification in that while both 
INTENT and INTRINSIC are unrelated to the primary independent measures, 
they correlate very strongly with each other. 

The Inverted-U Relationship Between Goal and Effort Expended 

It has been mentioned earlier in this chapter that there seems to 
be some indication that the relationship between goal level and the 
expenditure of effort is one that is characterized by peak values of 
effort corresponding to intermediate goal levels. For the purpose of 
further testing this possibility (as proposed in hypothesis 3) , analyses 
involving contrasts between goal levels 4 and 5 were undertaken. 
General linear models employing contrast codes for isolating the 
difference between the dependent variables corresponding to goal levels 


4 and 5 were employed for this purpose. It was found that as long as 
self-efficacy was not included in the models, the contrast was 
statistically significant (i.e., goal level 4 corresponded to a higher 
level of the dependent variable than goal level 5) for the dependent 
measure LAST OFFER for both the negotiation sessions. For the dependent 
measure TOTAL TIME, this drop approached significance only for the first 
of the two negotiation sessions (corresponding to the client AmTech) ; 
while for the dependent measure ROUNDS, no significant differences were 
found. With self-efficacy (and the interaction of self-efficacy with 
the contrast) included in the GLM model as additional independent 
variables, the drop in effort expenditure was evident only from the 
significantly higher LAST OFFER for goal level 4 (than the LAST OFFER 
for goal level 5) corresponding to the first negotiation session. For 
the second negotiation session, this difference approached significance. 
The other dependent variables did not yield significant differences. 

Thus, the results of the detailed analyses provide only partial 
support for the second hypothesis. 


This final chapter consists of concluding remarks regarding 
managerial implications of the observed results, limitations of the 
study, and directions for future research. 

Implications of Observed Results 

The conceptual model proposed in Chapter 3 identifies a number of 
significant interactive effects between externally assigned target 
performance levels, self-efficacy, and other constructs that affect 
motivation and performance. The experimental study confirms some of the 
key propositions of the model. 

A significant finding of the experimental study is that although 
self-efficacy does influence how individuals respond to externally 
assigned goals, the effect is only marginal at low goal levels or for 
very easy tasks. It is only when the task begins to increase in 
difficulty that the effect of self-efficacy is strongly felt. Indivi- 
duals with higher levels of self-efficacy respond more positively to 
increased levels of goal difficulty. They also exhibit higher 
subjective probabilities of task success. Furthermore, although each 
individual may be thought of as having a certain level of target 



performance which corresponds to a maximum level of actual performance, 
for those that are high in self -efficacy, this peak performance level is 
higher than that of those who are low in self-efficacy. 

Another important finding of this study is that providing indivi- 
duals with information about the probabilities of their success in the 
assigned task affects their personal expectancies either positively or 
negatively, depending upon the nature of the information supplied. That 
is, if a sales manager chooses to inform a salesperson that a given task 
will be a difficult one, then that information may detract from the 
salesperson's personal (i.e., independent) expectancy estimate and lower 
motivation for the task. Similarly, providing positive information 
(regarding the salesperson's high level of probability of achieving task 
success) should have a favorable impact on his/her motivation, 
especially when the task difficulty is actually high. Also, it was 
observed that individuals who are high in self-efficacy are likely to 
revise their expectancies upward to a greater extent (than those who are 
low in self-efficacy) when informed about the high level of probability 
of his/her achieving success at the given task. This finding thus has 
obvious implications for the manner in which a given task should be 
"framed" when it is assigned to an individual salesperson. 

Included among the limitations of the study is the lack of 
definitive support for two important propositions suggested by the 
theoretical framework. First, the results of the data analyses failed 
to clearly identify "peak" effort levels of the subjects. In almost all 
cases, the mean effort levels of the subjects are highest corresponding 
to the fourth goal level. However, the drop in effort expended when the 


externally assigned goal is raised to the fifth level is not 
statistically significant in several cases. The data as it exists 
suggest that the the values of effort expended level out (i,e., reach a 
plateau) at higher goal levels. In view of the evidence from prior 
studies, it seems likely that even higher levels of externally assigned 
goals than what was incorporated for this study could have resulted in 
significantly lower performance levels. Thus, in all probability, the 
problem lies with the calibration of the levels of the factor externally 
assign ed goal level . Secondly, the results seem to indicate that the 
effort expended by most subjects reach a maximum corresponding to the 
fourth goal level. It would have been possible to make a stronger 
statement about the impact of self-efficacy if individuals who were 
higher in self-efficacy seemed to expend peak levels of effort 
corresponding to higher levels of externally assigned goals than those 
who were low in self-efficacy. Since self-efficacy was measured (rather 
than manipulated) in this study, it is possible that the factor may not 
have had the benefit of as much variation as is required for purpose of 
producing such effects. 

Finally, from a theoretical standpoint, the experimental study did 
not include the examination of the dynamic aspects of goal setting. 

Future Research Directions 

The experimental study described in Chapter 4 addressed some 
significant aspects of the static issues related to goal setting (i.e., 
issues which do not take into account the effects of feedback and 
repetition) . Although the static aspects of goal setting have substan- 


tial value in the context of proving a theoretical base for future 
research, some of the more interesting problem situations in sales 
management involve the dynamic aspects of goal setting, in which 
feedback and repetitive performance goals are the norm, and where self- 
efficacies are more than likely to change over time. 

Future studies that explore the dynamic nature of self-efficacy 
and subsequent effects on motivation and performance would represent 
important contributions. In such contexts, experimental studies 
involving attempts at manipulating self-efficacies as opposed to 
measuring them, would constitute the preferred methodological format. 

An important property of the self-efficacy measure is that it is 
domain specific. Consequently, the necessity of a survey-based field 
study for the purpose of developing and validating a self-efficacy 
instrument applicable in an organizational setting cannot be over- 
emphasized. Although it would be relatively more difficult to 
administer, longitudinal studies that track the assigned tasks, the 
development of an individual salesperson's self-efficacy, intrinsic 
motivation, and performance over an extended period of time would 
contribute invaluable information in the domain of knowledge pertaining 
to motivation in the sales force. Validation of the propositions 
arising out of goal theory and self-efficacy studies within the context 
of several different actual sales tasks would constitute one of the 
major objectives of the future research directions indicated by this 
dissertation . 

The theoretical framework developed in this dissertation focuses 
specifically on sales related tasks and the performance goals discussed 


in the context of the framework refer exclusively to the sales 
situation. However, a number of marketing functions and activities are 
characterized by goals and motivated performance of individual task 
performers is an important issue in a majority of those areas. The 
basic tenets of goal theory could therefore be valuable in the explora- 
tion of the determinants of efficiency in a wide variety of areas 
related to marketing. Also, while this study has specifically focussed 
on the impact of externally assigned goals on performance at the 
individual level, it is conceivable that performance goals assigned at 
the group level (such as, those assigned for departments, organizations, 
etc.) influence overall performance as well. 


1. I am good at negotiating tasks. 

2. It is easy for me to get others to agree to my terms or see my 
point of view. 

3. I find it difficult to convince someone else, when that person's 
point of view conflicts with mine.* 

4. In any negotiation situation, I find it relatively easy to refuse 
an offer that I do not like. 

5. I find it very difficult to refuse an offer if the person I am 
negotiating with is very persistent.* 

6. It is easy for someone to take advantage of me when he (or she) is 
bargaining with me.* 

7. In most bargaining situations I am likely to come out as the 
"winner" when the deal is concluded. 

8. I can put pressure on someone I am bargaining with to get him or 

her closer to my terms. 

9. I do not possess the skills that are required to be a good 


A person has to be a really good negotiator to persuade me to 
accept his or her terms (or point of view) that I started out 
disagreeing with. 

11. My friends consider me to be a good negotiator. 

12. I cannot bring myself to apply pressure on another person in order 
to get a better deal for myself in a bargaining situation.* 

13. Because I am not a good bargainer, I am likely to do poorly in 
situations involving bargaining.* 

14. If a situation calls for negotiating, my friends would consider me 
to be the right person to deal with it. 



15. I usually come out better off in situations in which I have the 
opportunity to negotiate terms. 

16. My closest friends do not consider me to be a good negotiator.* 

17. I find it difficult to continue bargaining for long over any 
issue . 

18. I do noi give up easily in the course of negotiation until the 
conditions and terms that I am looking for have been met. 

19. In the course of a negotiation, I am good at "reading" an opponent 
with regard to what terms he or she is actually interested in. 

20. I can easily conceal my own minimum acceptable terms ans make 
offers and bids that will give me a substantially greater 

Note: Items marked with asterisks are reversed. 


1. I feel that I am a person of worth, at least on an equal basis with 
others . 

2. I feel that I have a number of good qualities. 

3. All in all, I am inclined to feel that I am a failure. 

4. I am able to do things as well as most other people. 

5. I feel that I do not have much to be proud of. 

6. I take a positive attitude toward myself. 

7. On the whole, I am satisfied with myself. 

8. I wish I could have more respect for myself. 

9. I certainly feel useless at times. 

10. At times I think I am no good at all.* 
Note: Items marked with asterisks are reversed. 




1. Personal Efficacy (E-I) (.75) .31 .12 .04 -.37 .19 

2. Interpersonal Control (E-I) (.77) .07 .35 -.28 .11 

3. Sociopolitical Control (E-I) (.81) -.24 -.50 -.03 

4. Machiavellanism (.68) .35 -.19 

5. Rotter's Locus of Control (I-E) (.70) -.32 

6. Marlowe-Crowne (.75) 

Note: 1) The reported reliabilities appear in parentheses. 

2) E-I refers to a scale orientation that measures the internality 
of perceived control; whereas I-E refers to a scale orientation 
that measures the externality of perceived control. 

3) n=110. 



The following sections illustrates how the intervening measures 
were collected in the course of the study. Although the specific 
examples listed here pertain to the negotiation task with AmTech only, 
the intervening measures collected for the other negotiation sessions 
were done in a very similar fashion. The exact questions and items that 
were presented to the subjects in the context of the interactive 
negotiation task on the computer are reproduced in the following 
sections . 

Subjec tive Difficulty of Goal (for the quota assigned jointly for 
AmTech & Consol) : This variable was measured by the following single 

Choose any number (between 1 and 7) that corresponds to your 
impression of the level of difficulty of the quotas that were 
assigned to you in the context of negotiating with AmTech and 
Consol : 

1. Extremely easy 

2. Fairly easy 

3. Somewhat easy 

4. Neither easy nor difficult 

5. Somewhat difficult 

6. Fairly difficult 

7. Extremely difficult 

Expectancy of Task Success (for the negotiation session with 
AmTech) : This variable was measured by the following single item. 



How do you rate the probability of your being able to achieve the 
quota that has been assigned to you with respect to your 
negotiations with AMTECH? Please type a whole number from 
through 100. 

Likelihood of Task Success (for the negotiation session with 
AmTech) : This variable was measured by the following single item. 

Choose any number (from 1 through 7) that corresponds to your 
estimate of how LIKELY you think it is that you will be able to 
achieve the sales quota that was assigned to you in the context of 
negotiation with AMTECH. 

1. NOT likely to achieve quota at all 

2. Extremely UNlikely to achieve quota 

3. Somewhat UNlikely to achieve quota 

4. Equally likely to achieve quota and not achieve quota 

5. Somewhat likely to achieve quota 

6. Extremely likely to achieve quota 

7. Certain of achieving quota 

Confidence Rating (associated with the expectancy and likelihood 
estimates for the negotiation task with AmTech) : This was measured by 

the following single item. 

Please indicate how CONFIDENT you were of your two preceding 
estimates of how well you are likely to do in the context of 
negotiating with AMTECH. That is we wish to know to what extent 
you were sure or unsure of the probability and likelihood 
estimates that you made just now. 

1. I was not confident in making the decisions at all 

2. I was very much lacking in confidence in making the estimates 

3. I was somewhat lacking in confidence in making the estimates 

4 . I neither lacked nor possessed confidence in making the 

5. I was somewhat confident about my estimates 

6. I was very confident about my estimates 

7. I was totally confident about my estimates 

Intrinsic Interest (corresponding to the assigned quota for the 
negotiation task with AmTech) : This was measured using a 7-point Likert 
format scale anchored by the extremities "Strongly AGREE" (corresponding 
to the point 1 on the scale) and "Strongly DISAGREE" (corresponding to 
the point 7) and consisting of the following three items. 


Please indicate your level of agreement with the following 
statements : 

1. "I think I will have a great deal of fun in trying to achieve 
the quota that was assigned to me for my negotiations with 
AmTech. " 

2. "I think that trying to achieve the quota for AmTech will be a 
very boring task." * 

3. "I feel challenged and motivated by the quota that was assigned 
to me for negotiations with AmTech." 

* The item marked with an asterisk was reversed before the scale score 
was computed. 

Intention to E xpend Effort (corresponding to the assigned quota 

for the negotiation task with AmTech) : This was measured using a 7- 

point Likert format scale anchored by the extremities "Strongly AGREE" 

(corresponding to the point 1 on the scale) and "Strongly DISAGREE" 

(corresponding to the point 7) and consisting of the following item. 

"I intend to work very hard in order to realize the quota that was 
assigned to me in the context of the negotiation with AMTECH." 

The intervening variables in the context of the negotiation task 
with CONSOL were measured similarly. 


Stage 1 

The computer program makes an initial offer on behalf of Banner (set 
at $15, 000) . 

Subject (types in response and/or) makes counteroffer. 

1(a) If the subject's offer is greater than a preset, maximum 
initial offer limit (set at $64,000 for Banner), then the 
program transmits a message 1 (from bank 7 of Banner's message 
file) which will alert the subject of the possibility of having 
entered an unusually large dollar amount, either on account of 
a typographical error or otherwise. The subject is requested 
to reconsider and submit a new counteroffer. 

1(b) If the subject makes a counteroffer that is less than Banner's 
initial offer, the computer program warns the subject of a 
possible typographical error, and offers him/her the option of 
changing it . 

1 (c) If the subject makes a counteroffer that is less than or equal 
to Banner's initial offer plus a certain set amount, known as 
tolerance (set at 0.06 x Banner's initial offer), then the 

x Refer to Appendix G for a sample of the messages used by the 
program to convey the client's responses to the subject. 



program issues an acceptance message (from bank 2) and returns 
the control of the terminal to the sales manager. 

1 (d) If the subject makes a counteroffer in the working range, 

(i.e., if the counteroffer is less than the maximum initial 
offer limit, but greater than Banner's previous offer plus 
tolerance), then the program issues a message from bank 4. 
Messages in this bank inform the subject of the client's 
inability to accept the counteroffer. Also, a new offer is 
made to the subject according to the formula: 

New offer = Banner's previous offer + {(Subject's offer - 
Banner's offer) * Concession slope} 

For Banner, the the concession slope is set at 0.007. 

Stage 2 

This second stage is reached only if the subject makes an initial 
counteroffer in the working range. In that case, the computer 
program provides the first revised offer on behalf of Banner in 
response to the subject's counteroffer. The program will then prompt 
the subject for a new counteroffer. 

2(a) If the subject's first revised counteroffer is greater than 
his/her previous counteroffer, Banner issues a message (from 
bank 5) that requests the subject not to move in the wrong 
direction in the context of the negotiations. The subject is 
requested to reconsider Banner's previous (revised) offer. 

2(b) If the subject's first revised counteroffer is exactly equal to 
his/her previous counteroffer, Banner's message informs the 
subject that the negotiations may proceed only if some 
concessions ar e made . The subject is requested to reconsider 
Banner's previous (revised) offer. A message from bank 6 is 


2(c) If the subject's first revised counteroffer is less than 

Banner's first revised offer, a warning message regarding a 
possible typographical error is issued and the subject is given 
an option to correct the entry. 

2(d) If the subject's first revised counteroffer is less than or 
equal to Banner's first revised offer plus tolerance, the 
counteroffer is accepted, and the acceptance message from bank 
2 is issued. 

2(e) If the subject's first revised counteroffer is not sufficiently 
less than his/her previous counteroffer, then Banner issues a 
message indicating the client's disapproval of the 
insignificance of the concession made (and a message from bank 
8 is issued) . In mathematical terms, this condition arises if 
the first revised counteroffer is greater than the previous 
offer minus an intolerance amount. The intolerance amount = 
previous offer * intolerance factor. Or, the condition is: 
Subject's counteroffer > (previous counteroffer - previous 
counteroffer * intolerance factor) . 
For Banner, the intolerance factor was set at 0.005. 

2(f) The remaining possibility involves a counteroffer made in the 
working range which is now defined as greater than Banner's 
previous offer plus tolerance, but less than or equal to the 
subject's prior counteroffer minus the intolerance amount. In 
this case, a second revised offer is made by Banner, using the 
same formula used earlier in Step 1 (d) and a fresh message from 
bank 4 is issued. 

Stage 2 as described in the preceding section is sufficiently general 
to represent the following stages. That is, for subsequent counter- 
offers, the conditional responses described in Stage 2 may be 
reiterated until the program terminates. The subject is informed 
beforehand that he/she may indicate an acceptance of Banner's offer 


at any time by typing an amount exactly equal to Banner's offer for 
the corresponding counteroffer. 

The algorithms used for the the negotiations with AmTech and Consol 
were very similar. The differences include the use of entirely 
different message modules (or message files) for each client. These 
modules differ from one another with respect to the language and 
sentence structures used but having similar meanings in terms of the 
implications of the messages for the corresponding message banks 
within modules. 

The different adjustable parameters, such as client's (i.e. the 
computer program's) initial offer, concession slope, etc., were 
changed from the settings used for Banner, but were kept identical 
for the two subsequent negotiations. 


Part of the Sample No. of Obs. Minimum Maximum Mean Std. Dev, 

Entire Sample 112 1.87 6.85 4.79 0.97 

1st quartile (Lowest) 

2nd quartile (Low) 

3rd quartile (High) 

4th quartile (Highest) 























A typical client was equipped with a number of messages for the 
purpose of communicating his/her own evaluation and responses to the 
subject. That is, the text file accessed by the computer program (to 
determine how a particular client should respond in a given situation) 
contained a variety of messages to suit the different actions that the 
program considered appropriate as responses to the subject's actions. 
The several different response situations may be categorized into a 
total of eight classes, where each of these situations actually 
represent a possible scenario that may result from the subject's choice 
of response. In other words, a client could be expected to confront a 
maximum of eight different situations which would require altogether 
different kinds of messages. Message banks corresponding to each of the 
eight different kinds of response situations were compiled so that each 
bank contained at least one message that could deal with the situation 
For instance, if the subject chose to accept the first offer made by the 
client, then a message from Bank number 1 was issued. This message 
essentially thanks the subject for complying with the initial offer and 
bids farewell. Because a given subject can confront this message only 
once, this bank contains a single message. Other banks contain more 
messages depending upon the how frequently the situation associated with 
that bank may arise. The following section contains samples of the 
different kinds of messages that were made available to the computer 
program for the purpose of issuing on behalf of one client. 



BANK 1 : Msg 1: We are glad that you found our offer acceptable. 

It was nice working with you. We are looking 
forward to working with you in the future. 

BANK 2 : Msg 2: Thank you. We will accept your offer. 

working with you. 'Bye! 

It was nice 

BANK 3 : Msg 3: The amount you have quoted is greater than what we 

can pay. Please consider our offer of 

Msg 4: It is unfortunate, but we cannot quote beyond a 

certain limit set by our company policies. We have 
made an offer as high as we possibly can. We hope 
will find our next offer of acceptable. Our offer 
is . . . 

Msg 5: As we have indicated before, there is no more room 
for us to quote any higher. It is quite impossible 
for us to make an offer higher than what we have 
already made. Once again, we are offering . . . 

BANK 4 : Msg 6: We are sorry, but we are unable to accept your 

offer. Please let us make you a new offer of . . 

Msg 7: Unfortunately, the amount you have quoted is still 
unacceptable. Please consider our new offer of . . 

Msg 8: Sorry, we have to refuse your offer again. Your 

quotation is too high in comparison to what we can 
pay. We are prepared to increase our offer to . . 

Msg 9: Thanks very much, for making this concession. We 
are still unable to evaluate your offer as 
acceptable. Do you think it will be possible for 
you to accept our offer of . . . 

Msg 10: We are sorry but we simply cannot offer you that 
amount. Please consider our offer of . . . 

BANK 5 : Msg 11: You have previously quoted us an amount that is 

lower than your last quoted amount. Please do not 
move away from a settlement. 

Msg 12: Once again, your preceding quote is HIGHER than a 
previously quoted price for your product. We are 
requesting you NOT to move in a direction that can 
jeopardize an agreement! 


Msg 13: If you feel that you must reevaluate the standards 
on which we have discussed prices so far, we will 
have lost the advantages from the progress we have 
made thus far. Please quote us an amount that is 
lower than your lowest quote till now. 

Msg 14: WE ARE SERIOUS! If you think you cannot proceed 
with these negotiations so that we may gradually 
converge on a settlement, we will be wasting our 
time. We insist that you quote lower than your 
lowest quoted amount thus far. 

Msg 15: Please quote an appropriate amount. If your 

quotation is higher than your lowest quoted amount 
till now, we will have no way of responding. We 
don't wish to have our negotiations deadlocked at 
this stage. Please quote an amount that is lower 
than your lowest quoted amount thus far. 

Msg 16: You MUST make a quotation that reflects a reduced 

price from your previous reasonable quote for us to 
be able to consider you at all. 

BANK 6 : Msg 17: Our continued negotiations are based on the 

understanding that both parties will make 
concessions in order to reach an agreement. If, we 
are to make any headway, you are requested to 
respond to our offers with counter-offers that 
reflect reasonable concessions. 

Msg 18: I realize that you do place a great deal of 

importance on the amount you have quoted. However, 
the only way we can make progress is if we can find 
ways to concede. We cannot accept your offer or 
make a counteroffer if you do not concede. 

Msg 19: Please realize that we are very serious about this: 
we CANNOT make any counteroffer whatsoever, unless 
you make some concession. 

Msg 20: We will be wasting your time and mine if if and 

when you refuse to make a concession. PLEASE make 
us an offer that is lower than this. 

Msg 21: We do have to tell you again that there is NO WAY 
that we can make you a new offer UNLESS you lower 
your prices by some amount when you make a new 
quote. Our system does not allow us to even review 
your new quote if it is not lower than your 
previous quote. 

Msg 22: Let us tell you in case you are overtly worried 
about your performance: We have negotiated with 
agents from IMC before. You won't be singled out 


for bad performance if you DO make the sale. 
Please provide us with some concession. Else, both 
you and I will have wasted our time. 

Msg 23: If you cannot concede, we will really fail to make 
any more progress than what we have accomplished 
thus far. Please believe me when I say that I 
cannot get an approval for submitting a quotation 
unless you lower your offer. 

Msg 24: We are serious about this! We cannot make a 
counteroffer unless your offer reflects a 

BANK 7 : Msg 25: You have quoted an unusually high amount. Please 

check for an error in typing and re-submit your 

Msg 26: Are you serious about the amount you have quoted? 
We don't know of anybody who would seriously make 
or even consider an offer as high as that. Please 
make us a sensible offer. 

Msg 27: You have quoted an impossible amount! Please check 

and resubmit your quote. You have to REALLY lower 

your offer before we can even begin to evaluate 

Msg 28: The amount you have quoted is prohibitively high! 
There is no way for us to evaluate an amount as 
high as that. We must ask you to lower your quote 
substantially for us to be able to respond. 

Msg 29: Please make a quotation that is not so high that we 
cannot even begin to evaluate it. You must lower 
your quote significantly for us to be able to 
continue negotiating. 

BANK 8 : Msg 30: We would really appreciate your concession if it 

was somewhat more generous than that. 

Msg 31: It is not that we do not appreciate your making 

this concession, . . . However, we believe that we 
could spend our time more efficiently if our offers 
and counteroffers reflect SIGNIFICANT concessions. 

Msg 32: We find your concession to be too little. I am 
sure that it is within your capabilities to let 
this negotiation proceed at a quicker pace. We are 
hopeful that you will now make us an offer that 
reflects a genuine concession. 


Msg 33: Thank you for conceding. But please realize that 
this offer that you have made is so close to your 
previous offers that it will not be worth our time 
to evaluate it separately. Please make an offer 
that amounts to a more generous concession. 

Msg 34: If you are serious about continuing negotiations 
with us, please lower your selling price by a 
reasonable amount. 

Msg 35: Please take this opportunity to make a realistic 
concession. We really cannot continue to waste 
your time and ours anymore. 


Alegra, J. A. (1983). 'Objective' and Perceived Task Characteristics as 
a Determinant of Reactions by Task Performers. Journal of 
Occupational Psychology , 56, 95-107. 

Anderson, CD.; Crowell, C.R.; Doman, M. ; & Howard, G.S. (1988). 

Performance Posting, Goal Setting, and Activity-Contingent 
Praise as Applied to a University Hockey Team. Journal of 
Applied Psychology , 73:1, 87-95. 

Anderson, R.E.; Hair, J.F.; & Bush, A.J. (1988). Professional Sales 
Management. . New York: McGraw-Hill Book Company. 

Andrews, F.M.; & Farris, G.F. (1972). Time Pressure and Performance of 
Scientists and Engineers: A Five-Year Panel Study. 
Organizational Behavior and Human Performance . 8, 185-200. 

Atkinson, J.W. (1958) . Toward an Experimental Analysis of Human 
Motivation in Terms of Motives, Expectancies, and 
Incentives. In J.W. Atkinson (Ed.), Motives in Fantasy, 
Action and Society , pp. 288-305. Princeton, NJ: Van 

Atkinson, J.W. (1964) . A Theory of Achievement Motivation. In J.W. 

Atkinson (Ed.), An Introduction to Achievement Motivation , 
pp. 240-268. Princeton, NJ: Van Nostrand. 

Atkinson, J.W. (1978) . The Mainsprings of Achievement -Oriented 
Activity. In J.W. Atkinson & J.O. Raynor (Eds.), 
Personality, Motivation, and Achievement , pp. 11-39. 
Washington, DC: Hemisphere. 

Atkinson, J.W.; & Feather, N.T. (Eds.) (1966). 
Motivation . New York: Wiley. 

A Theory of Achievement 

Atkinson, J.W.; & McClelland, D.C. (1948). The Projective Expression of 
Needs. II. The Effect of Different Intensities of the 
Hunger Drive on Thematic Apperception. Journal of 
Experimental Psychology , 38, 643-658. 

Atkinson, J.W.; & Reitman, W.R. (1958). Performance as a Function of 

Motive Strength and Expectancy of Goal-Attainment. In J.W. 



Atkinson (Ed.), Motives in Fantasy, Action, and Society , pp. 
278-287. Princeton, NJ: Van Nostrand. 

Bagozzi, R.p. (1978). Salesforce Performance and Satisfaction as a 
Function of Individual Difference, Interpersonal, and 
Situational Factors. Journal of Marketing Research , 15 
(November), 517-531. 

Bagozzi, R.P. (1980). Performance and Satisfaction in an Industrial 
Sales Force: An Examination of Their Antecedents and 
Simultaneity. Journal of Marketing , 44 (Spring), 65-77. 

Bandura, A. (1977a) . Social Learning Theory . Englewood Cliffs, NJ: 
Prentice-Hall . 

"Bandura, A. (1977b) . Self-Efficacy: Toward a Unifying Theory of 

Behavioral Change. Ps ychological Review , 84, 191-215. 

Bandura, A. (1982a) . Self-Efficacy Mechanism in Human Agency. American 
Psychologist , 37:2, 122-147. 

Bandura, A. (1982b) . The Self and Mechanisms of Agency. In J. Suls 
(Ed.), Psychological Perspectives on the Self (Vol. 1) , 
pp. 3-39. Hillsdale, NJ: Erlbaum. 

Bandura, A. (1984). Recycling Misconceptions of Perceived Self- 
Efficacy. Cognitive Therapy and Research , 8:3, 231-255. 

'•Bandura, A.; & Adams, N.E. (1977). Analysis of Self-Efficacy Theory of 
Behavioral Change. Cognitive Therapy and Research , 1:4, 

Bandura, A.; & Cervone, D. (1983). Self -Evaluative and Self-Efficacy 
Mechanisms Governing the Motivational Effects of Goal 
Systems . Journal of Personality and Social Psychology , 
45:5, 1017-1028. 

Bandura, A.; & Schunk, D.H. (1981). Cultivating Competence, Self- 
Efficacy, and Intrinsic Interest Through Proximal Self 
Motivation. Journal of Personality and Social Psychology , 
41:3, 586-598. 

Bandura, A.; & Simon, K.M. (1977). The Role of Proximal Intention in 
Self-Regulation of Refractory Behavior. Cognitive Therapy 
and Research , 1, 177-193. 

Bandura, A.; Adams, N.E.; & Beyer, J. (1977). Cognitive Processes 

Mediating Behavioral Change. Journal of Personality and 
Social Psychology , 25:3, 125-139. 

Bandura, A.; Adams, N.E.; Hardy, A.B.; & Howells, G.N. (1980). Tests of 
the Generality of the Self-Efficacy Theory. Cognitive 
Therapy and Research , 4:1, 39-66. 


Barling, J.; & Beattie, R. (1983). Self-Efficacy Beliefs and Sales 

Performance. Journal of Organizational Behavior Management , 
5:1, 41-51. 

Becker, L.J. (1978). Joint Effect of Feedback and Goal Setting on 
Performance: A Field Study of Residential Energy 
Conservation. Journal of Applied Psychology , 63:4, 428-433. 

Bernier, M. ; & Avard, J. (1986). Self-Efficacy , Outcome, and Attrition 
in a Weight-Reduction Program. Cognitive Therapy and 
Research , 10:3, 319-338. 

Betz, N.E.; & Hackett, G. (1981). The Relationship of Career-Related 
Self-Efficacy Expectations to Perceived Career Options in 
College Women and Men. Journal of Counselling Psychology , 
60, 257-265. 

Brucks, M. (1988) . Search Monitor: An Approach for Computer-Controlled 
Experiments Involving Consumer Information Search. Journal 
of Consumer Research , 15:1, 117-121. 

Campbell, D.J. (1982). Determinants of the Choice of Goal Difficulty 
Level: A Review of Situational and Personality Influences. 
Journal of Occupational Psychology , 55, 79-95. 

Campbell, D.J.; & Ilgen, D.R. (1976). Additive Effects of Task 

Difficulty and Goal Setting on Subsequent Task Performance. 
Journal of Applied Psychology , 61, 319-324. 

Campbell, J. P.; Dunnette, M.D.; Lawler, E.E., III; & Weick, K.E. (1970). 
Managerial Behavior, Performance, and Effectiveness . New 
York: McGraw-Hill. 

Campion, M.A.; & Lord, R.G. (1982). A Control-Systems Conceptualization 
of the Goal Setting and Changing Process. Organizational 
Behavior and Human Performance , 30, 265-287. 

Clement, S.(1987). The Self-Efficacy Expectations and Occupational 

Preferences of Females and Males. Journal of Occupational 
Psychology , 60, 257-265. 

Clopton, S.W. (1983). Alternative Research Designs and Methodologies 
For Buyer-Seller Negotiation Experiments. In W.R. Darden, 
K.B. Monroe & W.R Dillon (Eds.), Proceedings. 1983 AMA 
Winter Educators' Conference: Research Methods and Causal 
Modeling in Marketing , pp. 52-54. Chicago: American 
Marketing Association. 

Clopton, S.W. (1984). Seller and Buying Firm Factors Affecting 

Industrial Buyers' Negotiation Behavior and Outcomes. 
Journal of Marketing Research , 21 (February), 39-53. 

Clopton, S.W.; & Barksdale, H.C., Jr. (1987). Microcomputer Based 
Methods for Dyadic Interaction Research in Marketing. 


Journal of the Academy of Marketing Science , 15:2 (Summer), 

Condiotte, M.M.; & Lichtenstein, E. (1981). Self-Efficacy and Relapse 
in Smoking Cessation Programs. Journal of Consulting and 
Clinical Psychology , 49, 648-658. 

Cron, W.L.; Dubinsky, A. J.; & Michaels, R.E. (1988). The Influence of 
Career Stages on Components of Salesperson Motivation. 
Journal of Marketing , 52 (January), 78-92. 

Dachler, H.P; & Mobley, W.M. (1973). Construct Validation of an 
Instrumentality-Expectancy-Task-Goal Model of Work 
Motivation: Some Theoretical Boundary Conditions. Journal 
of Applied Psychology Monograph . 58:3, 397-418. 

Dalrymple, D.J. (1988) . Sales Management: Concept and Cases , 3 rd 
Edition. New York: Wiley. 

Deci, E.L. (1971). Effects of Externally Mediated Rewards on Intrinsic 
Motivation. Journal of Personality and Social Psychology . 
18:1, 105-115. 

DiClemente, C.C. (1981) . Self-Efficacy and Smoking Cessation 

Maintenance: A Preliminary Report. Cognitive Therapy and 
Research , 5, 175-187. 

DiClemente, C.C; Prochaska, J.O.; & Gibertini, M. (1985). Self- 
Efficacy and the Stages of Self-Change of Smoking. 
Cognitive Therapy and Research . 9:2, 181-200. 

Drucker, P.F. (1954). The Practice of Management . New York: Harper & 
Bros . 

Early, P.C.; Wojnaroski, P.; & Prest, W. (1987). Task Planning and 
Energy Expended: Exploration of How Goals Influence 
Performance. Journal of Applied Psychology , 72:1, 107-114. 

Eason, R.G. (1963). Relation Between Effort, Tension Level, Skill, and 
Performance Efficiency in a Perceptual Motor Task. 
Perceptual and Motor Skills , 16, 297-317. 

Eason, R.G.; & White, C.T. (1961). Muscular Tension, Effort, and 

Tracking Difficulty: Studies of Parameters Which Affect 
Tension Levels and Performance Efficiency. Perceptual and 
Motor Skills , 12, 331-372. 

Eastman, C . ; & Marzillier, J.S. (1984). Theoretical and Methodological 
Difficulties in Bandura's Self -Efficacy Theory. Cognitive 
Therapy and Research , 8:3, 213-229. 


Erez, M.; & Zidon, I. (1984). Effect of Goal Acceptance on the 

Relationship of Goal Difficulty to Performance. Journa l of 
Applied Psychology , 69:1, 69-78. 

Feldman, J.M.; & Lynch, J.G., Jr. (1988). Self Generated Validity and 

Other Effects of Measurement on Belief, Attitude, Intention, 
and Behavior. Journal of Applied Psychology , 7 3 (August) , 

Feltz, D.L. (1982). Path Analysis of the Causal Elements in Bandura ' s 
Theory of Self-Efficacy and an Anxiety-Based Model of 
Avoidance Behavior. Journal of Personality and Social 
Psychology , 42:4, 764-781. 

Feltz, D.L.; & Mungo, D.A. (1983). A Replication of the Path Analysis 
of the Causal Elements in Bandura 's Theory of Self-Ef f icacy 
and the Influence of Autonomic Perception. Journal of Sport 
Psychology , 5, 263-277. 

Feltz, D.L.; Landers, D.M.; & Raeder, U. (1979). Enhancing Self- 
Efficacy in a High-Avoidance Motor Task: A Comparison of 
Modeling Techniques. Journal of Sport Psychology , 1, 112- 

Frost, P.J.; & Mahoney, T.A. (1976). Goal Setting and the Task Process: 
I. An Interactive Influence on Individual Performance. 
Organizational Behavior and Human Performance , 17, 328-350. 

Futrell, C. (1981) . Sales Management: Behavior, Practice, and Cases . 
Hillsdale, IL: The Dryden Press. 

Garland, H. (1982). Goal Levels and Task Performance: A Compelling 

Replication of Some Compelling Results. Journal of Ap plied 
Psychology , 67, 245-248. 

Garland, H. (1983) . Influence of Ability, Assigned Goals, and Normative 
Information on Personal Goals and Performance: A Challenge 
to the Goal Attainability Assumption. Journal of Applied 
Psychology , 68:1, 20-30. 

Garland, H. (1984) . Relation of Effort-Performance Expectancy to 

Performance in Goal-Setting Experiments. Journal of Applied 
Psychology , 69:1, 79-84. 

Garland, H. (1985) . A Cognitive Mediation Theory of Task Goals and 

Human Performance. Motivation and Emotion , 9:4, 345-367. 

Gist, M.E. (1987). Self-Efficacy: Implications for Organizational 
Behavior and Human Resource Management. Academy of 
Management Review , 12:3, 472-485. 


Glynn, S.M.; & Ruderman, A.J. (1986). The Development and Validation of 
an Eating Self-Ef f icacy Scale. Cognitive Therapy and 
Research , 10:4, 403-420. 

Godding, P.R.; & Glasgow, R.E. (1985). Self-Efficacy and Outcome 

Expectations as Predictors of Controlled Smoking Status. 
Cognitive Therapy and Research , 9:5, 583-590. 

Griffin, C.E. (1928) . Sales Quota Systems. Michigan Business Studies , 
March, 1-43. 

Hackman, J.R.; & Lawler, E.E., III (1971). Employee Responses to Job 
Characteristics. Journal of Applied Psychology Monograph , 
55, 259-286. 

Hackman, J.R.; & Oldham, G.R. (1975). Development of the Job Diagnostic 
Survey. Journal of Applied Psychology , 60, 159-170. 

Hackman, J.R.; & Oldham, G.R. (1976). Motivation Through the Design of 
Work: Test of a Theory. Organizational Behavior and Human 
Performance , 16, 250-279. 

Hall, D.T.; & Hall, F.S. (1976). The Relationship Between Goals, 

Performance, Success, Self-image, and Involvement Under 
Different Organizational Climates. Journal of Vocational 
Behavior , 9, 267-268. 

Haring, A.; & Morris, M.L. (1968). Contests, Prizes, Awards for Sales 
Motivation . New York: Sales & Marketing Executives 
International . 

Hill, T.; Smith, N.D.; & Mann, M.F. (1987). Role of Efficacy 

Expectations in Predicting the Decision to Use Advanced 
Technologies: The Case of Computers. Journal of Applied 
Psychology , 72:2, 307-313. 

Hollenbeck, J.R.; & Klein, H.J. (1987). Goal Commitment and the Goal- 
Setting Process: Problems, Prospects, and Proposals for 
Future Research. Journal of Applied Psychology , 72:2, 212- 


Hollenbeck, J.R.; & Williams, C.R. (1987). Goal Importance, Self-Focus, 
and the Goal-Setting Process. Journal of Applied 
Psychology , 72:2, 204-211. 

Huber, V.L.; & Neale, M.A. (1987). Effects of Self- and Competitor 

Goals on Performance in an Interdependent Bargaining Task. 
Journal of Applied Psychology . 72:2, 197-203. 

Ingram, T.N.; & Bellenger, D.N. (1983). Personal and Organizational 
Variables: Their Relative Effect on Reward Valences of 
Industrial Salespeople. Journal of Marketing Research , 20 
(May), 198-205. 


Janz, T. (1982). Manipulating Subjective Expectancy Through Feedback: A 
Laboratory Study of the Expectancy-Performance Relationship. 
Journal of Applied Psychology , 67:4, 480-485. 

Jeffery, K.M. (1977) . The Effects of Sub-Goal Difficulty on Self- 
Motivated Persistence . Doctoral dissertation, Stanford 
University. Dissertation Abstracts International , 38:06 B, 
2940. (University Microfilms No. 77-25684). 

Johnson, H.W. (1976) . Sales Management: Opertations, Administration 
Marketing . Columbus, OH: Charles E. Merrill Publishing 
Company . 

Johnston, W.J.; & Bonoma, T.V. (1983). The Effect of Power Differences 
on the Outcome of Consumer Bargaining Situations. In 
T. Kinnear (Ed.), Advances in Consumer Research , Vol. 11, 
pp. 170-174. Ann Arbor, MI: Association for Consumer 
Research . 

Kahneman, D. (1973) . Attention and Effort . Englewood Cliffs, NJ: 
Prentice-Hall . 

Kelley, B. (1986). Recognition Reaps Rewards. Sales and Marketing 
Management , June, 100-104. 

Kohli, A.K. (1985). Some Unexplored Supervisory Behaviors and Their 
Influence on Salespeople's Role Clarity, Specific Self- 
Esteem, Job Satisfaction, and Motivation. Journal of 
Marketing Research , 22 (November), 424-433. 

Komorita, S.S.; & Barnes, M. (1969). Effects of Pressures to Reach 

Agreement in Bargaining. Journal of Personality and Social 
Psychology , 13:3, 245-252. 

Korman, A.K. (1971). Expectancies as Determinants of Performance. 
Journal of Applied Psychology , 55:3, 218-222. 

Latham, G.P.; & Locke, E.A. (1979). Goal-Setting: A Motivational 

Technique That Works. Organizational Dynamics , 8:2, 68-80. 

Lawler, E.E., III (1970). Job Attitudes and Employee Motivation: 

Theory, Research, and Practice. Personnel Psychology , 23, 

Lee, C. (1982) . Self-Efficacy as a Predictor of Performance in 

Competitive Gymnastics. Journal of Sport Psychology , 4, 

Lee, C. (1984) . Efficacy Expectations and Outcome Expectations as 
Predictors of Performance in a Snake Handling Task. 
Cognitive Therapy and Research , 8:5, 509-516. 


Lewin, K. (1938) . The Conceptual Representation and the Measurement of 
Psychological Forces . Durham, NC: Duke University Press. 

Lewin, K.; Dembo, T . ; Festinger, L.; & Sears, P.S. (1944). Level of 
Aspiration. In J. McV. Hunt (Ed.), Personality and the 
Behavior Disorders . Vol. 1, pp. 333-37 8. New York: Ronald 
Press . 

Lied, T.R.; & Pritchard, R.D. (1976). Relationships Between Personality 
Variables and Components of the Expectancy-Value Model . 
Journal of Applied Psychology , 61, 463-467. 

Locke, E.A. (1964) . The Relationships of Intentions to Motivation and 
Affect . Doctoral dissertation, Cornell University. 
Dissertation Abstracts International , 25, 6081. (University 
Microfilms No. 65-3729) . 

Locke, E.A. (1966). The Relation of Intention to Level of Performance. 
Journal of Applied Psychology , 69:2, 241-251. 

Locke, E.A. (1968). Toward a Theory of Task Motivation and Incentives. 
Organizational Behavior and Human Performance , 3, 157-189. 

Locke, E.A. (1978) . The Ubiquity of the Technique of Goal Setting in 

Theories of and Approaches to Employee Motivation. Academy 
of Management Review , 3, 594-601. 

Locke, E.A.; & Latham, G.P. (1984). Goal Setting: A Motivational 

Technology That Works ! Englewood Cliffs, NJ: Prentice-Hall. 

Locke, E.A.; Frederick, E . ; Lee, C . ; & Bobko, P. (1984). Effect of 
Self-Efficacy, Goals, and Task Strategies on Task 
Performance. Journal of Applied Psychology , 69:2, 241-251. 

Locke, E.A.; Latham, G.P.; & Erez, M. (1988). The Determinants of Goal 
Commitment. Academy of Management Review , 12:3, 472-485. 

Locke, E.A.; Shaw, K.N.; Saari, L.M.; & Latham, G.P. (1981). Goal 
Setting and Task Performance: 1969-1980. Psychological 
Bulletin , 90:1, 125-152. 

Locke, E.A; Mento, A.J.; & Katcher, B.L. (1978). The Interaction of 

Ability and Motivation in Performance: An Exploration of the 
Meaning of Moderators. Personnel Psychology , 31, 269-280. 

London, M.; & Oldham, G.R. (1976). Effects of Varying Goal Types and 

Incentive Systems on Performance and Satisfaction. Academy 
of Management Journal , 19, 537-546. 

Mace, C.A. (1935). Incentives: Some Experimental Studies . Report 72. 
London: Industrial Health Board. 


Manning, M.M.; & Wright, T.L. (1983). Self-Efficacy Expectancies, 

Outcome Expectancies, and the Persistence of Pain Control in 
Childbirth. Journal of Personality and Social Psychology , 
49, 421-431. 

Marzillier, J.; & Eastman, C. (1984). Continuing Problems With Self- 
Efficacy Theory: A Reply to Bandura . Cognitive Therapy and 
Research , 8:3, 257-262. 

Matsui, T.; & Tashitake, T. (1975). A Cross-Cultural Study of the 

Validity of Expectancy Theory of Work Motivation. Journal 
of Applied Psychology , 60, 263-265. 

Matsui, T.; Okada, A.; & Mizuguchi, R. (1981). Expectancy Theory 

ys Prediction of the Goal Theory Postulate, "The Harder the 
Goals, the Higher the Performance." Journal of Applied 
Psychology . 66:1, 54-58. 

McAlister, L.; Bazerman, M.H.; & Fader, P. (1986). Power and Goal 
Setting in Channel Negotiations. Journal of Marketing 
Research , 23 (August), 228-236. 

McAuley, E. (1985). Modeling and Self-Efficacy : A Test of Bandura's 
Model. Journal of Sport Psychology , 7, 283-295. 

McAuley, E.; & Gill, D. (1983). Reliability and Validity of the 

Physical Self-Efficacy Scale in a Competitive Sport Setting. 
Journal of Sport Psychology , 5, 410-418. 

McClelland, D.C.; & Atkinson, J.W. (1948). The Projective Expression of 
Needs. I. The Effect of Different Intensities of the Hunger 
Drive on Perception. Journal of Psychology , 25, 205-232. 

McClelland, D.C.; Atkinson, J.W.; Clark, R.A.; & Lowell, E.L. (1953). 

The Achievement Motive . New York: Appleton-Century-Crofts . 

McClelland, D.C.; Clark, R.A.; Roby, T.B.; & Atkinson, J.W. (1949). The 
Projective Expression of Needs. IV. The Effect of Need for 
Achievement on Thematic Apperception. Journal of 
Experimental Psychology ,, 39, 242-255. 

McGregor, D. (1954). The Human Side of Enterprise . New York: Harper & 
Row . 

Meier, S.; McCarthy, P.R.; & Schmeck, R.R. (1984). Validity of Self- 
Efficacy as a Predictor of Writing Performance. Cognitive 
Therapy and Research , 8:2, 107-120. 

Mento, A.J.; Cartledge, N.D.; & Locke, E.A. (1980). Maryland vs. 

Michigan vs. Minnesota: Another Look at the Relationship of 
Expectancy and Goal Difficulty to Task Performance. 
Organizational Behavior and Human Performance , 25, 419-440. 


Miller, M.L. (1979). Motivating the Sales Force. The Conference Board 
Information Bulletin , 64, 1-7. 

Mitchell, L.H. (1979) . Organizational Behavior. Annual Review of 
Psychology , 30, 243-281. 

Mitchell, L.H.; Smyser, CM.; & Weed, S.E. (1975). Locus of Control: 
Supervision and Work Satisfaction. Academy of Management 
Journal , 18, 623-631. 

Motowidlo, S.J.; Loehr, V.; & Dunnette, M.D. (1978). A Laboratory Study 
of the Effects of Goal Specificity on the Relationship 
Between Probability of Success and Performance. Journal of 
Ap plied Psychology , 63:2, 172-179. 

Naylor, J . C . ; & Ilgen, D.R. (1984). Goal Setting: A Theoretical 

Analysis of A Motivational Technology. In B. Staw & L.L. 
Cummings (Eds.), Research in Organizational Behavior, Vol.6 , 
pp. 95-140. Greenwich, CT : JAI Press. 

Neslin, S.A.; & Greenhalgh, L. (1983). Nash's Theory of Cooperative 
Games as a Predictor of the Outcomes of Buyer-Seller 
Negotiations: An Experiment in Media Purchasing. Journal of 
Marketing Research , 20 (November), 368-379. 

Odiorne, G.S. (1978). MBO: A Backward Glance. Business Horizons , 
(October) , 14-24. 

Oldham, G.R. (1975). The Impact of Supervisory Characteristics on Goal 
Acceptance. Academy of Management Journal , 18, 461-475. 

Oldham, G.R. (1976) . Organizational Choice and Some Correlates of 

Individuals' Expectancies. Decision Sciences , 7 (October), 

Oliver, R.L. (1974). Expectancy Theory Predictions of Salesmen's 

Performance. Journal of Marketing Research , 11, 243-253. 

Paulhus, D. (1983). Sphere-Specific Measures of Perceived Control. 

Journal of Personality and Social Psychology , 44:6, 1253- 

Pennington, A.L. (1968). Customer-Salesman Bargaining in Retail 

Transactions. Journal of Marketing Research , 5 (August), 

Peters, L.H. (1977). Cognitive models of Motivation, Expectancy Theory 
and Effort: An Analysis and Empirical Test. Organizational 
Behavior and Human Performance , 20 (October), 129-148. 

Phelps, D.M.; & Westing, J.W. (1968). Marketing Management . Homewood, 
IL: Richard D. Irwin. 


Pittman, T . S . ; Davey, M.E.; Alafat, K.A.; Wetherill, K.V.; & Kramer, 
N.A. (1980). Informational vs Controlling Rewards. 
Personality and Social Psychology Bulletin , 6, 228-233. 

Pittman, T . S . ; Emery, J.; & Boggiano, A.K. (1982). Intrinsic and 

Extrinsic Motivational Orientations: Reward-Induced Changes 
in Preference for Complexity. Journal of Personality and 
Social Psychology , 42:5, 789-797. 

Porter, L . W . ; & Lawler, E.E., III (1968). Managerial Attitudes and 
Performance . Homewood, IL: Irwin. 

Pritchard, R.D.; De Leo, P. J.; & Von Bergen, C.W., Jr. (1976). A Field 
Experimental Test of Expectancy-Value Incentive Motivation 
Techniques. Organizational Behavior and Human Performance , 
15, 335-406. 

Rollnick, S.; & Heather, N. (1982). The Application of Bandura ' s Self- 
Efficacy Theory to Abstinence-Oriented Alcoholism Treatment. 
Addictive Behaviors , 7, 243-250. 

Rosenberg, M. (1965) . Society and the Adolescent Self-image . 
Princeton, NJ: Princeton University Press. 

Ryan, T.A. (1958). Drives, Tasks, and the Initiation of Behavior. 
American Journal of Psychology , 71, 74-93. 

Sales and Marketing Management . (1986). Table III-6. February 17, p. 57. 

Schurr, P . H . ; & Ozanne, J.L. (1985). Influences on Exchange Processes: 
Buyers' Preconceptions of a Seller's Trustworthiness and 
Bargaining Toughness. Journal of Consumer Research , 11 
(March), 939-953. 

Schwab, D.P.; & Dyer, L.D. (1973). The Motivational Impact of a 

Compensation System on Employee Performance. Organizational 
Behavior and Human Performance , 9, 215-225. 

Segal, Z.V.; & Marshall, W.L. (1986). Discrepancies Between Self- 
Efficacy Predictions and Actual Performance in a Population 
of Rapists and Child Molesters. Cognitive Therapy and 
Research , 10:3, 363-376. 

Sheridan, J.E.; Slocum, J.W.; & Min, B. (1975). Motivational 

Determinants of Job Performance. Journal of Applied 
Psychology , 60, 119-121. 

Silber, E.; & Tipett, J. (1965). Self-Esteem: Clinical Assessment and 
Measurement Validation. Psychological Reports , 16, 1017- 


Sims, H.P., Jr.; & Szilagyi, A.D.; & McKemey, D.R. (1976). Antecedents 
of Work Related Expectancies. Academy of Management 
Journal , 19:4 (December), 547-559. 

Skinner, B.F. (1953) . Science and Human Behavior . New York: Macmillan. 

Skinner, B.F. (1971) . Beyond Freedom and Dignity . New York: Knopf. 

Skinner, B.F. (1974) . About Behaviorism . New York: Knopf. 

Spiro, R.L.; Perrault, W.D., Jr.; & Reynolds, F. (1977). The Personal 
Selling Process: A Critical Review and Model. Industrial 
Marketing Management ,. 5, 351-364. 

Stanton, W.J.; & Buskirk, R.H. (1978). Management of the Sales Force , 
5 th Edition. Homewood, IL: Richard D. Irwin. 

Stedry, A.C. (1960). Budget Control and Cost Behavior . Englewood 
Cliffs, NJ: Prentice-Hall. 

Steers, R.M.; & Porter, L.W. (1974). The Role of Task-Goal Attributes 

in Employee Performance. Psychological Bulletin , 81:7, 434- 

Stern, L.W. ; Sternthal, B.; & Craig, C.S. (1973). Managing Conflict in 
Distribution Channels: A Laboratory Study. Journal of 
Marketing Research , 10 (May), 169-179. 

Sujan, H. (1983) . Salesperson Motivation: Strategy Vs Effort . Doctoral 
dissertation, University of California, Los Angeles. 
Dissert ation Abstracts International , 44:12 A, 3792. 
(University Microfilms No. 84-07977) . 

Taylor, F.W. (1911) . Principles of Scientific Management . New York: 

Teas, R.K. (1981). An Empirical Test of Models of Salespersons' Job 
Expectancy and Instrumentality Perceptions. Journal of 
Marketing Research , 18 (May), 209-226. 

Teas, R.K.; & McElroy, J.C. (1986). Causal Attributions and Expectancy 
Estimates: A Framework for Understanding the Dynamics of 
Salesforce Motivation. Journal of Marketing , 50 (January), 

Templeton, J. (1986) . Carving Out True Recognition. Sales and 
Marketing Management , June, 106-114. 

Tolman, E.C. (1932) . Purposive Behavior in Animals and Men . New York: 

Tolman, E.C. (1955). Principles of Performance. Psychological Review . 
62, 315-326. 


Tubbs, M.E. (1986). Goal Setting: A Meta-Analytic Examination of the 
Empirical Evidence. Journal of Applied Psychology . 71:3, 

Tyagi, P.K. (1982). Perceived Organizational Climate and the Process of 
Salesperson Motivation. Journal of Marketing Research , 19 
(May), 240-254. 

Tyagi, P.K. (1985). Relative Importance of Key Job Dimensions and 
Leadership Behaviors in Motivating Salesperson Work 
Performance. Journal of Marketing , 49 (Summer), 76-86. 

Vallis, T.M.; & Bucher, B. (1986). Self-Efficacy as a Predictor of 
Behavioral Change: Interaction with Type of Training for 
Pain Tolerance. Cognitive Therapy and Research , 10:1, 79- 

Vroom, V.H. (1964). Work and Motivation . New York: Wiley. 

Vroom, V.H. (1966). Organizational Choice: A Study of Pre- and Post- 
Decision Processes. Organizational Behavior and Human 
Performance , 1, 212-225. 

Walker, O.C., Jr.; Churchill, G.A., Jr.; & Ford, N.M. (1977). 

Motivation and Performance in Industrial Selling: Recent 
Knowledge and Needed Research. Journal of Marketing 
Research , 14 (May), 156-158. 

Weinberg, R.S.; Gould, D.; & Jackson, A. (1979). Expectations and 

Performance: An Empirical Test of Bandura's Self-Ef f icacy 
Theory. Journal of Sport Psychology , 1, 320-331. 

Weinberg, R.S.; Yukelson, D, ; & Jackson, A. (1980). Effect of Public 
and Private Efficacy Expectations on Competitive 
Performance. Journal of Sport Psychology , 2, 340-349. 

Weitz, B.A. (1979) . A Critical Review of Personal Selling Research: The 
Need For A Contingency Approach. In G. Albaum & 
G. Churchill (Eds.), Critical Issues in Sales Management: 
State-of-the-Art and Future Research Needs . Eugene, OR: 
University of Oregon, College of Business Administration. 

Weitz, B.A. (1981). Effectiveness of Sales Interactions: A Contingency 
Framework. Journal of Marketing , 45 (Winter), 85-103. 

Wheeler, K.G. (1983). Comparisons of Self-Efficacy and Expectancy 

Models of Occupational Preferences for College Males and 
Females. Journal of Occupational Psychology , 56, 73-78. 

Yukl, G.A.; & Latham, G.P. (1978). Interrelationships Among Employee 

Participation, Individual Differences, Goal Difficulty, Goal 
Acceptance, Goal Instrumentality, and Performance. 
Personnel Psychology , 31, 305-323. 


Jhinuk Chowdhury was born in Calcutta, India, on February 13, 
1956. After completing high school education, he was admitted to the 
bachelor's degree program in mechanical engineering at Jadavpur 
University in Calcutta. After earning the bachelor's degree, he worked 
as an engineering superintendent in a manufacturing plant owned by Union 
Carbide India Limited in Calcutta. 

In 1984, he enrolled in the doctoral program in marketing at the 
College of Business Administration of the University of Florida in 
Gainesville, Florida. In the August 1989, he received and subsequently 
accepted an offer to join the faculty of the College of Business 
Administration of the University of North Texas in Denton, Texas, as an 
Assistant Professor. 


I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 

Barton A. Weitz, ChaifcT 
Professor of Marketing 

I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation f. or the degree of 
Doctor of Philosophy. 

Alan G. 

Professor of Marketing 

I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 


V^ 1 



John G. Lynch, Jr. ( . .-■' 
Associate Professor of 

I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 

/tA{itftlt< ^ • 


Richard E. Romano 
Assistant Professor of 

This dissertation was submitted to the Graduate Faculty of the 
Department of Marketing in the College of Business Administration and to 
the Graduate School and was accepted as partial fulfillment of the 
requirements for the degree of Doctor of Philosophy. 

August 1990 

Dean, Graduate School 

/ ! ) 

* -